Fermentation and food safety 2001 adams & nout

Fermentation andFood Safety

Editors

Martin R.Adams, MSc, PhDReader in Food Microbiology

School of Biomedical and Life SciencesUniversity of Surrey

Guildford, United Kingdom

M. J. Robert Nout, PhDAssociate Professor

Laboratory of Food MicrobiologyWageningen University

Wageningen, The Netherlands

AN ASPEN PUBLICATION®Aspen Publishers, Inc.

Gaithersburg, Maryland2001

The author has made every effort to ensure the accuracy of the information herein. However, appropriate infor-mation sources should be consulted, especially for new or unfamiliar procedures. It is the responsibility of everypractitioner to evaluate the appropriateness of a particular opinion in the context of actual clinical situations andwith due considerations to new developments. The author, editors, and the publisher cannot be held responsiblefor any typographical or other errors found in this book.

Aspen Publishers, Inc., is not affiliated with the American Society of Parenteral and Enteral Nutrition.

Library of Congress Cataloging-in-Publication Data

Fermentation and food safety/ Martin R. Adams and MJ. Robert Noutp. cm.

Includes bibliographical references and index.ISBN 0-8342-1843-7

1. Fermented foods. 2. Food industry and trade—Safety measures. I. Nout, M. J.Robert. II. Adams, Martin, R.

TP371.44.F472001660'.28448—dc21

00-051087

Copyright © 2001 by Aspen Publishers, Inc.A Walters Kluwer Company

www.aspenpublishers.comAll rights reserved.

Aspen Publishers, Inc., grants permission for photocopying for limited personal or internal use.This consent does not extend to other kinds of copying, such as copying for general distribution,

for advertising or promotional purposes, for creating new collective works, or for resale.For information, address Aspen Publishers, Inc., Permissions Department,

200 Orchard Ridge Drive, Suite 200, Gaithersburg, Maryland 20878.

Orders: (800) 638-8437Customer Service: (800) 234-1660

About Aspen Publishers • For more than 40 years, Aspen has been a leading professional publisherin a variety of disciplines. Aspen's vast information resources are available in both print and electronicformats. We are committed to providing the highest quality information available in the most appropri-ate format for our customers. Visit Aspen's Internet site for more information resources, directories,articles, and a searchable version of Aspen's full catalog, including the most recent publications:vnvw.aspenpublishers.com

Aspen Publishers, Inc. • The hallmark of quality in publishingMember of the worldwide Wolters Kluwer group.

Editorial Services: Sharyn L. Vitrano/Timothy SniffinLibrary of Congress Catalog Card Number: 00-051087

ISBN: 0-8342-1843-7

Printed in the United States of America

1 2 3 4 5

Contributors

Martin R. Adams, MSc, PhDReader in Food MicrobiologySchool of Biomedical and Life SciencesUniversity of SurreyGuildford, United Kingdom

A. AsanteScientistFood Safety UnitWorld Heath OrganizationGeneva, Switzerland

R.R. Beumer, PhDAssociate ProfessorLaboratory of Food MicrobiologyWageningen University Research CenterWageningen, The Netherlands

Leon Brimer, PhD, DScAssociate ProfessorDepartment of Pharmacology and

PathobiologyLaboratory of ToxicologyRoyal Veterinary and Agricultural

UniversityCopenhagen, Denmark

Michael J. Carter, BA, PhDReader in VirologySchool of Biomedical and Life SciencesUniversity of SurreyGuildford, United Kingdom

Wilhelm H. Hozapfel, PhDDirector and ProfessorInstitute of Biotechnology and Molecular

BiologyFederal Research Centre for NutritionKarlsruhe, Germany

Maurice O. Moss, PhD, ARCS, DICVisiting ReaderSchool of Biomedical and Life SciencesUniversity of SurreyGuildford, United Kingdom

Yasmine Motarjemi, PhDScientistFood Safety ProgrammeWorld Health OrganizationGeneva, Switzerland

M.J. Robert Nout, PhDAssociate ProfessorLaboratory of Food MicrobiologyWageningen UniversityWageningen, The Netherlands

Arthur C. Ouwehand, PhDSenior ScientistDepartment of Biochemistry and Food

ChemistryUniversity of TurkuTurku, Finland

Seppo J. Salminen, PhDProfessorDegree Programme on Health SciencesDepartment of Biochemistry and Food

ChemistryUniversity of TurkuTurku, Finland

M. Hortensia Silla-Santos, PhDSenior ScientistInstitute de Agroquimica y Tecnologia de

AlimentosConsejo Superior de Investigaciones

CientificasValencia, Spain

Mike Taylor, BVMS, PhD, MRCVS, CBiol,MIBiol

Head of Parasitology and EcotoxicologyVeterinary Laboratories AgencyNew Haw, United Kingdom

T. Verrips, PhDChief ScientistUnilever Research LaboratoryVlaardingen, The Netherlands

Atte J. von Wright, PhDProfessorInstitute of Applied BiotechnologyUniversity of KuopioKuopio, Finland

Foreword

In every part of the world, people wage a con-stant battle against food contamination and theresulting food-borne diseases and food wastage.Efforts to reduce the devastating consequencesof food contamination started long before writ-ten records. Besides cooking, smoking, andsimple sun drying, fermentation is one of theoldest technologies used for food preservation.Over the centuries, it has evolved and has beenrefined and diversified. Today, a large variety offoods are derived from this technology, which isused in households, small-scale food industries,and large-scale enterprises. Foods so producedform a major part of the human diet all over theworld but only a few people are aware of themultitude of fermented products and their im-portance in the human diet. In fact, all cultureshave in the course of their development learnedthe technique to preserve some of their foods byfermentation. However, the safety of fermentedfoods is a concern everywhere.

In the past, traditional fermentation technolo-gies were based on experiences accumulated byconsecutive generations of food producers, as aresult of trial and error. Only relatively recentlyhave science and technology started to contrib-ute to a better understanding of the underlyingprinciples of the fermentation process and of theessential requirements to ensure the safety aswell as nutritional and sensory quality of fer-mented foods. Since the days of Louis Pasteur,who pointed to the importance of hygiene in re-lation to fermentation, it is known that this tech-nology is easily influenced by various factors

during processing, and if not applied correctly,the safety and/or quality of the final product maybe jeopardized. As a matter of fact, causes andoutbreaks of food-borne illness have been tracedback to fermented food, in spite of the generalineffectiveness of food-borne disease surveil-lance programs in most countries.

Fermentation is also of economic importancein areas or for populations where preservationtechnologies such as cold storage (refrigeration)or hot-holding cannot be used for lack of re-sources or facilities. In such situations, fermen-tation may be considered an affordable technol-ogy, which — if applied correctly — results in thesafe preservation of foods, including comple-mentary foods for infants. Particularly in devel-oping countries, as a result of poor hygiene andincorrect application of fermentation, comple-mentary foods are often contaminated withpathogens and subsequently are a major cause ofinfant diarrhea and associated malnutrition.

Against this background, the World HealthOrganization (WHO), jointly with the Food andAgriculture Organization (FAO), organized in1995 a workshop to assess fermentation as ahousehold technology for improving foodsafety.1 This workshop was the first of its kind,highlighting the critical points in the fermenta-

!Fermentation: Assessment and Research. Reportof a Joint FAOAVHO Workshop, Pretoria, South Af-rica, 1 1-15 December 1995. WHO Consultations andWorkshops: WHO/FNU/FOS/96.1

tion process to ensure the safety of the resultingproducts, in line with the Hazard Analysis andCritical Control Point (HACCP) system. In away. this book is a result of this workshop. BothAspen Publishers and the two editors, MJ. Rob-ert Nout and Martin R. Adams, deserve applausefor this initiative and the unique approach theyhave adopted: the book focuses on food safety inall its aspects and is largely hazard based, whichhelps to identify those areas where knowledge islacking but needed for a satisfactory risk assess-ment to be made. This is in contrast to the exist-ing literature on fermented foods, which is gen-erally confined to descriptions of the product(s)

and the microbiology/biochemistry of their pro-duction.

In the interest of public health and food secu-rity, I wish this book a large and interested read-ership and for fermentation to result in safe, nu-tritionally adequate, and superbly tasting foodswith long shelf lives.

Fritz Kdfer stein, DVM9 PhDFormer Director, Programme of Food

Safety and Food Aid, WHO, Geneva;Current Distinguished Visiting Scientist,

Joint Institute for Food Safety and AppliedNutrition, Washington, DC

Preface

Fermented foods enjoy worldwide popularityas attractive, wholesome and nutritious compo-nents of our diet. They are produced on an enor-mous scale employing a huge variety of ingredi-ents and manufacturing techniques. Whethertraditional home-made foods, or high-tech prod-ucts derived from genetically modified organ-isms, the safety of the consumer remains of fore-most importance.

Fermented foods have always been generallyregarded as safe, but this reputation has been se-riously threatened in recent years by incidentssuch as outbreaks of illness caused by pathogensin soft cheeses and fermented meats, chroniccyanide intoxications from poorly processedcassava tubers, and mycotoxins in fermented ce-real foods. In addition, modern techniques of ge-netic engineering and biotechnology, which of-fer considerable opportunities in the area offermented food production, have also raisedsafety concerns among consumers. It is neces-sary therefore to have concrete guidelines on theconditions which lead to safe products and tohave a realistic view about what "guaranteedsafety" means. The massive impact of HACCPas a systematic approach to ensuring food safetyhas been widely apparent in recent years and thetechnique is as applicable to food fermentationas to any other food processing operation. Atpresent however, the literature on fermentedfoods has no focus on safety, but is mainly de-scriptive, concentrating on microorganisms re-sponsible for fermentation and on the biochemi-cal changes occurring in the food.

This book aims to integrate modern conceptsof safety assurance with the sometimes very tra-ditional environment of the production and dis-tribution of fermented foods. In particular, wehave taken a largely hazard-based approachrather than one centered on the different com-modities used. Introductory chapters aim to pro-vide a broad understanding of the nature of fer-mented foods, their production, distribution, anduse by consumers, and also discuss the generalfeatures of fermentation processes that contrib-ute to the product's overall safety and HACCP.For the bulk of the book, we have sought chap-ters which describe the principal individual haz-ards, both chemical and microbiological, and tryto provide some guidance on how these might becontrolled in food fermentations. These hazardsare discussed from the point of view of their se-verity and incidence, how they get into the food,which foods are specifically at risk, and what, ifany, are the conditions that remove or inactivatethese hazards. In many cases there is a dearth ofpublished material specifically on fermentedfoods, and contributors have used data obtainedin slightly different contexts to give some guid-ance. This exercise has proved useful in high-lighting where information is lacking and identi-fying areas where more research is desperatelyneeded.

It is hoped that the book will serve as a sourceof reference to support and help improve theproduction of safe fermented foods at all scales(household preparation up to large-scale indus-trial plants) and using all major food groups.

iii This page has been reformatted by Knovel to provide easier navigation.

Contents

Contributors ............................................................................................. vii

Foreword ................................................................................................. ix

Preface .................................................................................................... xi

1. Fermented Foods and Their Production ....................................... 1 Fermentation and Food Safety ....................................................................... 1 Food Fermentation Components .................................................................... 4 Diversity of Fermented Foods ......................................................................... 13 Process Unit Operations ................................................................................. 15 Process Conditions ......................................................................................... 17 Appendix 1-A: Flow Diagrams for Selected Food Fermentations .................. 20

2. Why Fermented Foods Can Be Safe ............................................. 39 Introduction ...................................................................................................... 39 Physical Processing ........................................................................................ 39 Microbial Activity .............................................................................................. 41 Overall Significance of Different Antimicrobial Factors .................................. 48

3. An Introduction to the Hazard Analysis and Critical Control Point (HACCP) System and Its Application to Fermented Foods ............................................................................ 53 Introduction ...................................................................................................... 53 Historical Development ................................................................................... 53 The Need for Change ...................................................................................... 54 What Is the HACCP System? ......................................................................... 55 Benefits of the HACCP System ...................................................................... 56 Areas of Application ........................................................................................ 56

iv Contents

This page has been reformatted by Knovel to provide easier navigation.

The HACCP System in Food Hygiene ............................................................ 57 Application of the HACCP Approach to Food Preparation ............................. 58 Application of the HACCP System to Gari ...................................................... 60 Appendix 3-A: HACCP Principles and Guidelines for Its Application ............. 67

4. Chemical Hazards and Their Control: Endogenous Compounds ..................................................................................... 71 Introduction ...................................................................................................... 71 Toxic and Antinutritional Glycosides in Food and Feed ................................. 71 Risks Associated with the Occurrence of Toxic Glycosides in Different

Commodities .......................................................................................... 77 Variation in Toxin Concentration among Varieties and Cultivars: The

Influence of Traditional Domestication and Modern Breeding .............. 83 Removal of Toxins through Processing .......................................................... 84 Conclusion ....................................................................................................... 88

5. Chemical Hazards and Their Control: Toxins .............................. 101 Microbial Toxins .............................................................................................. 101 Aflatoxins ......................................................................................................... 102 Ochratoxin A .................................................................................................... 105 Patulin .............................................................................................................. 107 Fusarium Toxins .............................................................................................. 108 Alternaria Toxins ............................................................................................. 111 Bacterial and Algal Toxins ............................................................................... 112

6. Toxic Nitrogen Compounds Produced during Processing: Biogenic Amines, Ethyl Carbamides, Nitrosamines .................... 119 Introduction ...................................................................................................... 119 Biogenic Amines .............................................................................................. 119 Ethyl Carbamides ............................................................................................ 126 Nitrosamines .................................................................................................... 129

7. Microbiological Hazards and Their Control: Bacteria ................. 141 Introduction ...................................................................................................... 141 Outbreaks Related to Fermented Products .................................................... 141 Bacterial Pathogens of Concern in Fermented Products ............................... 142

Contents v


Presence of Pathogens in Raw Materials ....................................................... 146 Effect of Fermentation Processes on the Survival of Bacterial

Pathogens .............................................................................................. 147 The Control of Microbial Hazards ................................................................... 148 HACCPs .......................................................................................................... 150 Validation of the Sausage Fermentation Process for the Control of

Pathogens .............................................................................................. 152

8. Microbiological Hazards and Their Control: Viruses ................... 159 Introduction ...................................................................................................... 159 Food-Borne Viruses ........................................................................................ 160 Potentially Food-Borne Viruses ...................................................................... 163 Assessment of Virus Risk ............................................................................... 168

9. Microbiological Hazards and Their Control: Parasites ................ 175 Introduction ...................................................................................................... 175 Nematodes ...................................................................................................... 175 Cestodes (Tapeworms) ................................................................................... 187 Trematodes (Flukes) ....................................................................................... 192 Protozoa .......................................................................................................... 200 Conclusion ....................................................................................................... 211

10. Biotechnology and Food Safety: Benefits of Genetic Modifications ................................................................................... 219 Introduction ...................................................................................................... 219 Supply Chain of Fermented Foods and Its Importance to Ensure Safe

Products ................................................................................................. 221 Two Sets of Examples of Safety Evaluation of Fermented Foods for

Which Recombinant DNA Is Used ......................................................... 229 Conclusion ....................................................................................................... 236

11. Safety Assessment of Probiotics and Starters ............................ 239 Safety Aspects for Probiotics and Biotechnology ........................................... 239 Safety of Current Probiotics ............................................................................ 242 Relevance of a Sound Taxonomic Basis ........................................................ 242 Transfer of Antibiotic Resistance .................................................................... 246

vi Contents


Genetically Modified Organisms ..................................................................... 246 Conclusion ....................................................................................................... 247

12. Practical Applications: Prospects and Pitfalls ............................. 253 Introduction ...................................................................................................... 253 Benefits of Lactic Acid Fermentation .............................................................. 254 Pitfalls of Fermentation .................................................................................... 255 Importance of Food Fermentation in Public Health ........................................ 257 Practical Intervention to Enhance Safety of Fermented Foods ...................... 259 Research Needs .............................................................................................. 271 Conclusion ....................................................................................................... 271

List of Sources ....................................................................................... 275

Index ....................................................................................................... 279

CHAPTER 1

Fermented Foods andTheir Production

MJ. Robert Nout

FERMENTATION AND FOOD SAFETY

Fermentation

Fermentation is one of the oldest methods offood processing. Bread, beer, wine, and cheeseoriginated long before Christ. Although modernfood technology has contributed to the present-day high standard of quality and hygiene of fer-mented foods, the principles of the age-old pro-cesses have hardly changed. In industrializedsocieties, a variety of fermented foods are verypopular with consumers because of their attractiveflavor and their nutritional value (Figure 1-1).

In tropical developing regions, fermentation isone of the main options for processing foods. Inthe absence of facilities for home refrigeration,freezing, or home canning, it serves as an afford-able and manageable technique for food preserva-tion. Fermentation can also increase the safety offoods by removing their natural toxic components,or by preventing the growth of disease-causingmicrobes. It imparts attractive flavor and nutritivevalue to many products. Fermentation is an attrac-tive technique because it is low cost and low tech-nology and it can be easily carried out at the house-hold level, often in combination with simplemethods such as salting, sun drying, or heating(e.g., boiling, steaming, frying).

Contrary to unwanted spoilage or toxin pro-duction, fermentation is regarded as a desirableeffect of microbial activity in foods. The mi-crobes that may be involved include molds

(mycelial fungi), yeasts (unicellular fungi), andbacteria. Examples of food fermentations andthe microbes responsible for the desired changeswill be presented in this chapter.

In general, the desirable effect of microbialactivity may be caused by its biochemical activ-ity. Microbial enzymes breaking down carbohy-drates, lipids, proteins, and other food compo-nents can improve food digestion in the humangastrointestinal tract and thus increase nutrientuptake. Several bacteria excrete B vitamins intofood. As a result of their growth and metabo-lism, substances of microbial origin are found inthe fermented food, including organic acids,alcohols, aldehydes, esters, and many others.These may have a profound effect on the qualityof the fermented product. For instance, lacticand acetic acids produced by lactic acid bacteria(LAB) have an inhibitory effect on spoilage bac-teria in sourdough bread and yogurt, and the pro-duction of ethanol and carbon dioxide deter-mines the acceptability of bread, beer, and wine(ethanol disappears from bread during the bak-ing process).

In addition to enzymes and metabolites, mi-crobial growth causes increased amounts of mi-crobial cell mass. This may be of nutritional andaromatic interest in yeast extract, for instance.The presence of living microbial cells such as innonpasteurized yogurt may well have advanta-geous effects on the intestinal microflora and,indirectly, on human health.

Nr.

123456789

10111213141516171819

Name of product

quarkyogurtsauerkrautcultured milk ("karnemelk")treated black olivesgouda cheeseraw fermented sausagesyeast-leavened breadlager beerwhite winesherrylager beergruyere cheesepumpernickel breadmixed sourdough breadcamembert cheeseraw fermented sausagesoy saucetempeh

Major ingredient(s)

cows' milkcows' milkwhite cabbagecows' milkolivescows' milkpork and/or beef meatwheat flourbarley, hopsgrapesgrapesbarley, hopscows' milkrye flourrye + wheat flourcows' milkpork and/or beef meatsoybeans and wheatsoybeans

Functional microorganisms

lactic acid bacteria (LAB)LABLABLABLABLABLAByeasts (Y)YYYYLAB + propionibacteriaLAB + YLAB + YLAB + moldsLAB + moldsMolds + LAB + YLAB + Molds + Y

Figure 1-1 Fermented foods representing different types of fermentations and raw materials

Food Safety

In its widest sense, the safety of food must beachieved through safe production, storage, andhandling in order to avoid food-borne illnessessuch as food intoxication, infectious diseases, or

other detrimental effects. In principle, such ill-nesses can be caused by agents of biological,chemical, or physical nature, as exemplified inTable 1-1.

In this book, the fermentation of food will beviewed in relation to these safety aspects. The

Table 1-1 Some Examples of Threats to Consumer Safety

Illness

Enteric infections

Liver cancerRespiratory failureHypertension

Cyanide intoxicationIntoxications

HypertensionAllergies

Injuries (cuts, perforations)

Specific Examples

Bacteria (Salmonella spp.)Viruses (Rotavirus)Parasites (Cryptosporidium spp.)Mycotoxin-producing fungi (Aspergillus flavus)Toxin-forming bacteria (Clostridium botulinum)Biogenic amine-forming bacteria (Enterobacteriaceae)

Cyanogenic glycosides (linamarin) in cassava tubersPesticidesVeterinary drugsHeavy metalsBiogenic amines (e.g., in fish)PreservativesColorants

GlassMetal

Type of Causative Agents

Pathogenic microorganisms

Toxigenic microorganisms

PhytotoxinsEnvironmental contaminants

Toxic metabolitesFood additives and ingredients

Foreign matter

Nature

Biological

Chemical

Physical

book begins with a closer look at fermentation asa food processing technology. Next, the intrinsicsafety of fermented foods as well as the prin-ciples of the hazard analysis critical controlpoint (HACCP) system are discussed, followedby several examples of hazards that are potentialthreats to consumer safety.

The use of genetically modified ingredients ormicroorganisms in food fermentations, as wellas the use of microorganisms as probiotics infermented foods, are aspects that require a sys-tematic and critical assessment of their safety.

In the final synthesizing chapter, the HACCPapproach is used to illustrate and compare someof the critical processing steps that affect thesafety of fermented foods.

FOOD FERMENTATION COMPONENTS

Food Ingredients

Food ingredients include the raw foods cho-sen for fermentation which can be of plant oranimal origin. These foods contain a variety ofnutrients for the consumer but also some of thesenutrients will be required for the microbialgrowth and metabolism during the fermentationprocess. In order to enable microbial activity,sufficient water must be present. Consequently,water must be added in case the other ingredi-ents are too dry. For several reasons, such astaste or preservation, miscellaneous substancesmay be added to the ingredient mix. These willbe discussed in more detail below.

Food Groups

• Plant origin: Fermented foods of plant ori-gin are derived from a variety of raw mate-rials of different chemical composition andbiophysical properties. Tuberous rootssuch as potatoes and cereals and tree cropslike breadfruit have a relatively high starchcontent. Legumes and oil seeds generallyhave a high protein content. Green veg-etables, carrots, beets, tomatoes, olives, cu-cumbers, okra, and forage crops for animalfeed silages have a high moisture content.Fruits contain high concentrations of re-ducing sugars.

Most fermentative preservations of veg-etables and cereals are due to the action of LAB,often in combination with yeasts. But other bac-teria, such as Bacillus spp., or mycelial fungi,such as Rhizopus and Aspergillus spp., areequally important in the fermentation of le-gumes and oil seeds.

• Animal origin: Foods of animal origin thatare fermented are mainly milk, meat, fish,and seafood. These are all quite perishable,and fermentation has long been an effectivemethod for prolonging the shelf life ofthese valuable nutrients.

Milk originating from cows, buffaloes, sheep,and, occasionally, other animals has a highmoisture content; contains proteins, minerals,and vitamins; and has a neutral pH. A rapidacidification by lactic acid bacterial fermenta-tion to pH values of less than 4.5 strongly inhib-its the survival and growth of spoilage-causingor disease-associated bacteria. Milk containsample amounts of the fermentable carbohydratelactose, which is an essential ingredient to en-able this fermentation. Due to the reduced pH, aswell as bacterial glycocalix, the viscosity in-creases and various liquid fermented milk prod-ucts are obtained, ranging from fluid culturedmilk to stringy viscous Scandinavian milks,mixed sour-alcoholic fizzy fluids, and gel-typeproducts such as yogurt. Milk is also rather volu-minous; nomadic tribes developed methods tocurdle milk and separate the coagulated caseinfrom the residual liquid (whey). The coagulaterepresents only approximately 10% of the origi-nal milk volume. Fermentation by LAB, and oc-casional propionibacteria, yeasts, and molds, re-sults in a variety of cheeses.

Meat (cow, pork, goat, etc.) contains very lowlevels of fermentable carbohydrates, thus posinglimitations to lactic acid fermentation. Many ofthe fermented meats derive their long shelf lifefrom a combination of preservative effects.Acidification by LAB is usually boosted by add-ing sugar, and added salt, nitrate, or nitrite, aswell as some extent of dehydration, are majoraspects of the preservation of fermented sau-sages. It is interesting to note that most fer-

mented sausages have been prepared from rawmeat. In order to avoid the risk of raw meat-borne food infections by parasites, it is recom-mended to freeze the meat prior to processing.

Fish and seafood (e.g., shrimp) pose similarrestrictions to bacterial fermentation: They con-tain only low quantities of fermentable carbohy-drates. In practice, this has led to two majortypes of fish products preserved by fermenta-tion. The first is a mixture of fish and salt thatresults in liquid protein hydrolysate after severalmonths of fermentation. Halotolerant bacteriaand yeasts may play a role in flavor formation,but the high salt content is responsible for thepreservation. The second type of fish product isa mixture of fish, little salt, and cooked starchyfood (e.g., rice or cassava), the latter providingfermentable carbohydrates; these products arepreserved mainly by lactic acid fermentation.

Nutrients Required by Microorganisms

• Carbon and nitrogen: Most microorgan-isms require some form of organic carbon.In natural raw materials, this is foundmainly in carbohydrates (e.g., monomers,dimers, etc., and a number of polymericfood-storage and structure-giving polysac-charides such as starch, pectins, hemicellu-lose, and cellulose). Molds and certain bac-teria (e.g., Bacillus spp.) are goodproducers of extracellular carbohydrate-de-grading enzymes that can release ferment-able mono- and oligomeric substances. Ac-counts of the response of yeasts to sourcesof carbon, nitrogen, phosphorus, and sulfurare available.34'41 Yeasts and LAB areknown to be poor converters of polysaccha-rides and pentoses. Although the presenceof chitin and acacia gum was shown to in-crease the rate of yeast growth and fermen-tation,31 these compounds were not me-tabolized as nutrients.

In addition to fermentable saccharides, othernutrients required for cell growth and metabo-lism include inorganic (e.g., NH4+) or organicsources of nitrogen (e.g., urea, amino acids, andpeptides, but rarely extracellular proteins),

phosporus (e.g., inorganic phosphate or phos-phate esters), and sulfur (e.g., sulfate, sulfite,methionine, glutathione). Fungi usually do notrequire additional nutrients in food fermenta-tion. Optimum carbon/nitrogen ratios for growthare 10-100. For use in industrial fermentations,it was shown that R. oryzae could grow wellwith only starch and nitrogen salts; the additionof vegetable juice further stimulated growth.37

• Minerals: Iron, magnesium, potassium,sodium, and calcium are normally requiredfor cell growth.41 Of additional interest isthe effect of Ca2+ ions, which was reportedto increase the ethanol tolerance ofSaccha-romyces cerevisiae.24 Manganese plays animportant role in LAB; deficiencies maylead to fermentation failures.

• Vitamins and other growth factors: Themost common growth factors for yeasts arebiotin, pantothenic acid, inositol, thiamine,nicotinic acid, and pyridoxine.41 Riboflavinand folic acid are synthesized by all yeasts,but are required by some bacteria. Fungalrequirements for additional vitamins in thefood environment seem to be negligible.On the contrary, they may contribute to thenutritive value of fermented foods by vita-min synthesis.

Water

Water is essential for microbial growth andmetabolism. The extent to which water is avail-able for biological metabolism is expressed aswater activity (Aw) or, occasionally, as water po-tential.15 Aw is the more commonly used termi-nology and is defined by the ratio of equilibriumwater vapor pressure of the food and of pure wa-ter, at a defined temperature such as 20 0C. Awranges from O to 1. Most microorganisms re-quire Aw to be greater than 0.70, with optimumgreater than 0.99.

Added Food Ingredients

A variety of added ingredients affect the ac-tivity of microorganisms and thus can be used toregulate the rate and extent of fermentation. Saltand sugar are well known for their antimicrobial

effect at high concentrations. Salt levels greaterthan 15% w/w and sugar concentrations greaterthan 25% w/w reduce the Aw considerably; inaddition, NaCl has a specific inhibitory influ-ence caused by Na+ ions. Herbs, spices, and soforth may contain inhibitory proteins, organicacids, essential oils, pigments, resins, phenoliccompounds, caffeine, and so on. On the otherhand, some spices are excellent sources of man-ganese, which is essential for LAB.

Microorganisms

Three groups of microorganisms are used infood fermentation, namely bacteria, yeasts, andmolds. Table 1-2 illustrates some prominentspecies of microorganisms, some of the food fer-mentations in which they are of importance, andtheir function.

How do microorganisms enter the fermenta-tion? Different scenarios of increasing complex-ity can be distinguished,25 as will be explained inthe following sections.

Natural Fermentation in Raw Substrate

Most raw foods of animal or plant origin con-tain a variety of microorganisms that arrived bychance or that have an ecological associationwith them, based on preharvest growing condi-tions. Food processing activities can also addcertain microorganisms to food. If food is al-lowed to ferment without prior heating, most ofthese microorganisms can multiply. However,their opportunities will be restricted by theirability to grow and compete in the food, and byexternal conditions. This usually results in a suc-cession of predominant microorganisms, finallystabilizing in a fermented product that contains amixed microbial population dominated by mi-croorganisms that are particularly suited to thephysicochemical conditions prevailing in the fi-nal product. This type of fermentation is exem-plified by sauerkraut, which is shredded andsalted cabbage that is fermented by LAB.11

Drawbacks of natural fermentations are that theytake a relatively long time to complete, and theoutcome is always a surprise.

Use of Traditional Mixed StarterCultures in Raw or PreheatedSubstrate

The drawbacks of natural fermentation can bereduced when a large quantity of microorgan-isms that occur in the final product are added.These may be expected to bring about a morerapid domination and a more predictable qualityof the fermented product. An example of thisapproach is the traditional fermentation of sour-dough, a mixture of cereal flour (wheat or rye)with water of dough consistency.12 A smallquantity of previously fermented sourdough ismixed into the new dough, and this practice canbe successfully carried out for many years toachieve dependable fermentations.

Traditional mixed-culture starters are also ap-plied in the fermentation of precooked ingredi-ents. Examples are the Indonesian fermentedsoybean food tempeh (also known as tempe\ 18

as well as African alcoholic beverages such asGhanaian pito made from sorghum.9 Fortempeh, a mixed culture of molds (Rhizopus andMucor spp.) especially grown for this purposeon plant leaves (Hibiscus tileaceus) is used.27

The fermentation starter for pito is an inocula-tion belt, a woven rope that is suspended in thefermenting beer and on which the predominatingyeasts (Saccharomyces spp.) are present as asediment. The fermentation of a next batch ofpito beer is started by immersing the inoculationbelt into the sugary liquid made from sorghum,which is a common cereal in tropical climates.

Use of Pure Cultures (Single or Mixed)in Preheated Substrate

With increasing scale of production, more so-phisticated technical facilities, and higher in-vestment and operational risks, the use of labo-ratory-selected and precultured starter culturesbecomes a necessity. Because equipment foraseptic processing at a large production scale isextremely expensive, common procedures in-clude pure culture maintenance and propagationunder aseptic conditions (e.g., sterile laboratoryand pilot fermentors). At production scale, thefood ingredient to be fermented is preheated in

Table 1-2 Selected Examples of Microorganisms and their Function in the Fermentation of Foods

Importance of Fermentation forProduct Characteristics

Contribute to flavor, shelf life, structure, andconsistency by the production of lactic acid,acetaldehyde, diacetyl, and polysaccharides

Produce ethanol, CO, and flavor

Form proteolytic and saccharolytic enzymes,enabling solubilization and flavor production

Fermented Foods

Yogurt

Alcoholic beverages(beers, wines)

Soy sauce

Species

Lactic acid bacteria: Lactobacillus delbruckiissp. bulgaricus and Streptococcussalivarius ssp. thermophilus

Saccharomyces cerevisiae

Aspergillus sojae

Microbial Group

Bacteria

Yeasts

Molds

order to minimize the level of microbial contami-nation; subsequently, the pure culture starter(single or mixed cultures) is added and fermenta-tion takes place at the highest practicable and af-fordable level of hygienic protection.

Use of Pure Cultures in SterilizedSubstrate

For the propagation of starter cultures, or inother situations requiring the absolute absenceof microbial contamination, food ingredients orotherwise suitable nutrients are sterilized andkept in sterile confinement. Pure cultures ofstarter microorganisms are added using aseptictechniques. Not only is this an expensive tech-nique, but it is also unnecessary in production-scale food fermentation. Moreover, severe heattreatments required to achieve sterility can harmheat-sensitive nutrients for both the microorgan-ism and the consumer.

Enzymes

The importance of enzymes in fermentationprocesses lies in their ability to degrade complexsubstrates. Some examples are in koji fermenta-

tion, where A. oryzae enzymes degrade starch,protein, and cell wall components, and intempeh production, where Rhizopus spp. en-zymes degrade cell wall components, protein,lipids, phytic acid, and oligosaccharides. Pro-teases produced by LAB are important degrad-ing proteins in fermented milk products. Someenzymes involved in the degradation of complexsubstrates include amylolytic, proteolytic, Ii-polytic, and cell wall degrading enzymes.

Amylolytic Enzymes

Starch is composed of two polysaccharides,amylose and amylopectin, both consisting ofglucose units only (Figure 1-2). Amylose is a lin-ear glycan in which the sugar residues are con-nected by a- 1,4 bonds; amylopectin is abranched glycan in which glucose residues in thebackbone and the side chains are a- 1,4 linked.The side chains are attached by a- 1,6 linkages.

Starch is degraded by several amylases work-ing simultaneously, a-amylases hydrolyzea- 1,4 glucosidic linkages. Iso-amylase is adebranching enzyme; it hydrolyzes a- 1,6 gluco-sidic linkages. Amyloglucosidase liberatessingle glucose units from the nonreducing end,

Figure 1-2 Amylopectin molecule and its degradation by several amylases. Adapted with permission fromUhlig, H. (1991). Enzyme arbeiten fur uns. Technische enzyme und ihre Anwendung. Miinchen, Wien: CarlHanser Verlag.43

Glucose Maltose

8-Amylase

Amyloglucosidase

a-Amylase

iso-amylase orpullulanase

whereas (3-amylase liberates maltose units.These enzymes belong to the class of hydro-lases. In industrial-scale production of a-amy-lases, the enzyme is derived from the pancreaticgland of swine and cattle and from microbialcultures. Microorganisms that produce amylasesare B. subtilis and A. oryzae. B. licheniformisproduces a heat-stable amylase that can be usedat 100-1 10 0C. Gelatinization of starch by heat-ing enhances enzyme catalysis. Thus, the swol-len gelatinized starch substrate is degraded 300times faster by bacterial amylase and 105 timesfaster by fungal amylase than the native,unswollen, nongelatinized starch.

Proteases

Proteases are classified according to theirsource (e.g., animal, plant, microbial), their cata-lytic action (e.g., endopeptidase or exopeptidase),and the nature of the catalytic site. Endopeptidasesare the proteases that are used most commonly infood processing, but in some cases, they are ac-companied by exopeptidases. Endopeptidasescleave polypeptide chains at particularly suscep-tible peptide bonds distributed along the chain,whereas exopeptidases hydrolyze one amino acidfrom the end of the chain. The four major classesof endopeptidases are carboxyl proteases (i.e., as-partic protease), cysteine proteases, serine pro-teases, and metallo proteases (Table 1-3). As thenames imply, carboxyl, cysteine, and serine pro-teases have carboxyl, cysteine, and serine sidechains, respectively, as essential parts of the cata-lytic site. The carboxyl proteases generally havemaximum catalytic activity at acid pH. The serineproteases have maximum activity at alkaline pH;the closely related cysteine proteases usually showmaximum activity at more neutral pH values. Themetallo proteases contain an essential metal atom,usually Zn2+, and have an optimum activity at neu-tral pH. Ca2+ stabilizes these enzymes and strongchelating agents (such as ethylenediamine-tetraacetic acid, or EDTA) inhibit them.

Enzymes Degrading Cell WallComponents

In order to understand how enzymes affectplant cell walls, the structure of the walls first

must be reviewed. Microscopic investigationshave revealed that plant cell walls can be dividedinto three layers: the middle lamella, primarywall, and secondary wall (Figure 1-3). Themiddle lamella acts as an intercellular bindingsubstance and is mainly composed of pectin.Secondary cell walls contain less pectin but con-tain some lignin. Primary cell walls consist ofcellulose fibers called microfibrils embedded ina matrix of pectins, hemicelluloses, and proteins(Figures 1-3 and 1 )̂. Pectin is the major bind-ing component of the cell wall, and its degrada-tion by pectolytic enzymes will cause fruit orvegetables to become soft. This is the first stepof enzymatic degradation of the cell wall.

• Pectolytic enzymes: The basic structure ofpectins is a linear chain of oc-linked mol-ecules of pyranosyl D-galacturonic acid.Varying proportions of carboxylic groupscan be present as methyl esters. The esteri-fication is usually by methanol, in whichcase the pectin is said to be methylated.When less than 50% of the carboxyl groupsare methylated, the pectin is referred to aslow methoxyl pectin; when more than 50%of the carboxyl groups are methylated, thepectin is referred to as high methoxyl pec-tin. Pectin situated in the middle lamella isremoved from the cell wall relatively easilyand is most easily degraded by appropriateenzymes. In contrast, enzymes do not eas-ily degrade the pectin within primary andsecondary cell walls. Pectinases are definedand classified on the basis of their actiontoward the galacturonan part of the pectinmolecules. Two main groups are distin-guished, pectin esterases and pectindepolymerases.

• Hemicellulases: Hemicelluloses arepolysaccharides that are extracted fromplant cell walls by strong alkali. They arecomposed of four major substances:arabinans, galactans, xyloglucans, and xy-lans. As an example, arabinans are de-graded by the enzymes, arabinanases.Arabinans are branched polysaccharideswith a backbone of a- 1,5 linked L-arabinose

Table 1-3 Overview of Proteases

Optimum StabilitypH Range

6.5-6.0

3.0-5.0

4.5-6.5

7.5-9.56.0-8.0

5.07.07.0-8.03.0-6.03.8-6.5

pH-Optimum

6.0-7.02.09.0

7.0-8.07.0-8.07.0-8.0

7.0-1 1 .06.0-9.0

3.0-4.05.5-7.56.0-9.53.5-4.5

5.0

Source

Stomach lining of calvesGastric lining of swine or bovinePancreas

Tropical melon tree (Carica papaya)Pineapple (fruit and stalk)Figs (Ficus carica)

e.g., Bacillus subtilise.g., Bacillus thermoproteolyticusStreptomyces griseus

Aspergillus oryzaeAspergillus oryzaeAspergillus oryzaeMucor pusillusRhizopus chinensis

Type

Carboxyl proteaseCarboxyl protease

Cysteine proteaseCysteine proteaseCysteine protease

Serine proteaseMetallo protease

Carboxyl proteaseMetallo proteaseSerine proteaseCarboxyl proteaseCarboxyl protease

Name

Proteases of animal originChymosinPepsinPancreatic protease*Proteases of plant originPapainBromelainFicinBacterial proteasesAlkaline proteaseNeutral proteasePronaseFungal proteasesAcid proteaseNeutral proteaseAlkaline proteaseProteaseProtease

*A mixture of trypsin, chymotrypsin, and various peptidases with amylase and lipase as accompanying enzymes.

Primary wall

Middle lamella

Secondary wall

Figure 1-3 Schematic representation of the structure of plant cell walls. Reproduced with permission fromVoragen, A. G. J., Van den Broek, L. A. M. (1991). Fruit juices. Biotechnological Innovations in Food Process-ing, pp. 187-210. Edited by Biotol Team. Oxford, UK: Butterworth-Heinemann.44

units (Figure 1-5). To approximately everythird arabinose molecule, additional arabi-nose units are attached by a- 1,2- or a- 1,3linkages. This can produce complex struc-tures. There are two types of arabinanases:

arabinosidase (arabinofuranosidase) andendo-arabinanase. Arabinosidase can besubdivided into two forms, A and B. The Bform degrades branched arabinan to a lin-ear chain by splitting off terminal a- 1,3 or

Figure 1—4 Schematic representation of the primary cell wall. PGA and RG are part of pectin. Reproduced withpermission from Carpita, N. C., Gibeaut, D. M. (1993). Structural models of primary cell walls in floweringplants: consistency of molecular structure with the physical properties of the walls during growth. The PlantJournal*, 1-30.7

RG I witharabinogaiactanside-chains

PGA junctionzoneXyloglucan

cellulose fibril

Figure 1-5 The structure of arabinans. Adapted with permission from Beldman, G., Schols, H. A., Pitson, S. M.,Searle-van Leeuwen, M. J. F., Voragen, A. G. J. (1997). Arabinans and arabinan degrading enzymes. In Advancesin Macromolecular Carbohydrate Research. Vol. 1, pp. 1-64. Edited by RJ. Sturgeon. London: JAI Press.1

a- 1,2 linked side chains. At the same time,this enzyme slowly and sequentially breaksthe a- 1,5 linkages at the nonreducing endof linear arabinan. Endo-arabinanase hy-drolyzes linear arabinan in a random fash-ion, producing oligomers of shorterlengths. Arabinosidase A degrades thearabinan oligomers to monomers. Arabina-nases occur in some plants and microorgan-isms. Fungal arabinanases have an optimumpH of approximately 4.0; bacterialarabinanases have an optimum pH of ap-proximately 5 . 0-6 . 0 .

• Cellulases: Cellulose is the best known ofall plant cell wall polysaccharides. It is par-ticularly abundant in secondary cell wallsand accounts for 20-30% of the total drymass. Cellulose is a linear chain of P- 1,4linked glucose units. In cellulose, these p-1,4 glucan chains aggregate by hydrogenbonds to rigid flat structures called fibrils(Figure 1—4). The degree of polymerizationcan be very high, up to 14,000 in secondarycell walls. The rigid structure makes cellu-lose very resistant to degradation by en-zymes. Cellulase (often called the cellulasecomplex) is a multienzyme complex sys-tem composed of several enzymes: endo-

glucanase, exo-glucanase, cellobiose hy-drolase, and cellobiase.

Lipases

Lipases degrade triglycerides. They only acton an aqueous-lipid interface such as a micelle.Generally, the enzymes catalyze the hydrolysisof triglycerides to produce fatty acids and glyc-erol, but there are also specific enzymes thatcatalyze the hydrolysis of monoacylglycerides,phospholipids, and esters of sterols. Generally,the mode of action of lipases results in fatty ac-ids being preferentially hydrolyzed from 1- and 3-positions of triglycerides so as to leave 2-substi-tuted monoacylglycerides. Microbial lipases mayalso catalyze the hydrolysis in all three positions.The composition of the fatty acid (i.e., length,stereoconformation, and degree of saturation) af-fects the specificity and speed of the lipases.

Phytases

Phytases catalyze the hydrolytic removal ofphosphate groups from phytic acid (Figure 1-6),a substance that occurs widely in cereals and le-gumes and that is known to limit thebioavailability of macro- and micronutrients.The ability to degrade phytate occurs widelyamong molds (A.ficuum, R. oligosporus), yeasts

Figure 1-6 Structure of phytic acid

(Candida krusei, Schwanniomyces castelli), andLAB (Pediococcus pentosaceus, Lactobacillusagilis, Lb. confusus). Because the optimum pH ofmost phytases is approximately 5.5, the phytatedegradation can be less effective if a lactic acidfermentation has resulted in low pH values.

Tannases

Tannases catalyze the hydrolysis of polyphe-nols (tannins). Fungal enzymes (e.g., from A.niger) are used in the food industry to degradetannins. Tannin removal results in reduced tur-bidity in tea-based beverages. It also benefits thenutrient bioavailability in tannin-containing fer-mented cereal products.2

DIVERSITY OF FERMENTED FOODS

There is a wide variety of fermented foodsworldwide. Various excellent reference bookson fermented foods are available,5'40'45 as aresource books on industrial aspects of food bio-technology, including food fermentation.23'33'39

Mention was made already of the various foodingredients that can be fermented, as well as themicroorganisms and enzymes that are used infermentations.

In the strict sense, fermentation refers to aform of anaerobic energy metabolism. In thecontext of fermented foods, however, the micro-bial growth and metabolism can take place under

aerobic conditions as well. For example, moldfermentations require oxygen to facilitate theirgrowth and enzyme production.

An aspect of interest is the physical state un-der which the fermentation takes place. This canbe in the form of liquid in which the microorgan-isms are suspended while a relatively simplemixing device is used to ensure the homogeneityof microorganisms, substrates, and products. Inthese liquid fermentations, or LFs, the continu-ous phase is the liquid, and the control of tem-perature and levels of dissolved oxygen can beachieved with immersion coolers or heaters, andaeration. Quite a different situation occurs in aheap of cooked soybeans in which homogeneousgrowth of strictly aerobic molds is required.Whereas the particulate matter (i.e., soybeans)contains sufficient water to allow microbialgrowth, water is not the continuous phase; gas(i.e., air) is.This physical state is referred to assolid-state fermentation, or SSF. Because gas isa poor conductor of heat and mass, SSFs have atendency to develop gradients of temperatureand gas composition. Control of homogeneityrequires more complex measurement and mix-ing systems as compared to LF. In practice, in-termediate situations such as shredded veg-etables or particles such as olives, which arefermented while immersed in brine, can be en-countered. These can be considered as immer-sion LF because the continuous phase is liquidbut contains immersed particles. Obviously,these do not always behave like LF.

Table 1-4 illustrates selected examples offood fermentations representing different foodgroups, functional microorganisms and en-zymes, oxygen requirements, physical states,and present levels of industrialization. As men-tioned earlier, food ingredients of plant as wellas animal origin are used to prepare fermentedfoods. Of the functional microorganisms, LABplay the predominant role in the prolongation ofshelf life because of the antimicrobial effect oftheir acids. In addition, yeasts are often found asminority companions of LAB; they sometimescontribute to shelf life by scavenging residualassimilable carbon sources. This is the case, forexample, in fermented pickles. The major func-

Table 1-4 Examples of Food Fermentation Processes

References

30

16

26,28

3

10

36

42

8

Industrialization*

1-2

1-4

1-3

1-4

1-4

1-4

1-4

1-4

PhysicalState

Solid

Solid

Solid

Immersedin liquid

Liquid

Solid

Liquid

Solid

OxygenRequirement

Aerobic

Anaerobic

Aerobic

Anaerobic

Anaerobic

Anaerobic

Anaerobic

Aerobic

Enzymic Modificationsof The Product

Degradation of starchand cell walls

Degradation of starchand maltose

Degradation of protein,cell walls, and lipids

Degradation of lipids

Degradation ofproteins

Degradation ofproteins and lipids

Microorganisms andMajor Products*

Lactic acid bacteriaforming lactic andacetic acid

Lactic acid bacteria andyeasts forming lacticand acetic acid, andcarbon dioxide

Molds forming myceliumand enzymes

Lactic acid bacteriaforming lactic acid

Yeasts (and occasion-ally lactic acidbacteria) formingethanol and flavor

Micrococcaceae, lacticacid bacteria (andoccasionally molds)forming organic acidsand flavor andstabilizing meat color

Lactic acid bacteriaforming lactic acid,acetaldehyde, anddiacetyl and providingstructure

Lactic acid bacteria andmolds forming acidity,release enzymes forripening and flavordevelopment

Mode ofConsumption

Cooked paste

Bread

Fried snacks orcooked sidedish

Vegetable dishor relish

Alcoholicbeverage

Hearty foodingredient orsnack

Breakfast ordessert

Sandwichesand snacks

FermentedFood

Agbeli mawe

Sourdough

Tempeh

Sauerkraut

Wine

Raw driedsausages

Yogurt

Soft cheeses

Food Group

Starchy crops:cassava

Cereals: wheatand rye

Legumes:soybean

Vegetables:cabbage

Fruits: grape

Meats: pork andbeef

Milk: cow,buffalo

Milk: cow,sheep, goat

*Mention is made only of functional groups of microorganisms.f1 , home-produced; 2, village style; 3, national market; 4, international market.

tion of yeasts in yeast-fermented products istheir copious production of carbon dioxide, suchas gas in bread and beer and ethanol in alcoholicbeverages. The functionality of molds is mainlyin their production of enzymes that degradepolymeric components such as cell wallpolysaccharides, proteins, and lipids. This canhave considerable consequences for the texture,flavor, and nutritional value of mold- fermentedfoods. Depending on the functional microorgan-isms that should be propagated, the incubation en-vironment (i.e., availability of oxygen for strictaerobes, anaerobic conditions when mold growthis undesirable) and physical state can be selectedand optimized. It is of interest to note that, in con-trast to predominantly LFs that are common in thepharmaceutical industry, a considerable numberof food products are fermented by SSF.

PROCESS UNIT OPERATIONS

The manufacture of fermented foods is orga-nized in a sequence of activities called unit op-erations.22 Irrespective of the scale at which theprocess takes place, the following types of unitoperations can be distinguished.

Physical Operations

Transport

Transport is one of the most important unitoperations. It has the purpose of transferring in-gredients (i.e., transport of mass) to the desiredlocalities and/or equipment. Its aim is also to as-sist in heating and cooling (i.e., transport ofheat). There is a wide variety of materials andmethods that can be used for transport. Thechoice depends on the type of product, scale ofproduction, economic considerations, and localconditions.

Grading and/or Sorting

Grading and/or sorting have the purpose ofachieving homogeneity of size, color, maturity,hardness, and so forth. At the same time, itemsthat are spoiled, infected, or otherwise deterio-rated are removed. From a food safety point of

view, grading and sorting are important toolsthat can be used to optimize the quality of inputs.

Cleaning and/or Washing

Cleaning and/or washing are carried out to re-move dirt, dust, insects, agricultural residues,and so forth. Depending on the type of ingredi-ents, dry or wet cleaning can be chosen. In cerealprocessing, for example, dry cleaning of wheatprior to flour milling is performed by a combina-tion of aspiration and sieving. But, in maize pro-cessing, maize kernels can be washed in waterprior to wet milling. Water is becoming everscarcer and expensive, causing increased inter-est in dry operations.

Physical Separations

Dehulling, peeling, trimming, and other sepa-rations are aimed at obtaining the desired ana-tomical parts of plant or animal tissue while re-moving the undesired ones. Examples are theremoval of poorly digestible seedcoats fromsoybeans by dehulling, the removal of the cortexof the cassava root because of its bitter taste andpossible toxicity, and the trimming of fat fromred meat.

Moisture Adjustment

Moisture adjustment is required to achieve,for example, the desired consistency and edibil-ity of dry seeds and grains. The need for a suffi-ciently high Aw for microbial metabolism wasmentioned earlier. Water can be added as an in-gredient and mixed; soaking or steeping in wateris also a common method to increase the mois-ture content of ingredients. Especially duringlonger periods of soaking at favorable tempera-tures, microbial activity can take place. Severalcereal fermentation processes combine soakingwith fermentation.

Size Reduction

Size reduction of particulate matter is re-quired to obtain meal or flour from seeds, to ob-tain pulp from tubers or fuits, or to prepare a ho-mogeneous slurry of meat and other ingredientsfor sausage making. Size reduction is performedby cutting, grinding, impact hammering, and so

forth using a wide variety of equipment that hasbeen developed to suit the processing of specificraw materials.

Mixing

Mixing has the purpose of obtaining homoge-neity of mass and heat. Mixing of mass is themost common type of mixing and is exemplifiedby the mixing of ingredients to obtain a homoge-neous product. Mixing at a small scale is rela-tively easy and can be carried out with simplekitchen utensils. The larger the scale of opera-tion, the more complicated mixing becomes.Mixing of dry components can be achieved us-ing specific mixing equipment such as tumblersor augur-type mixers, but it is also feasible tocombine mixing with other operations such asgrinding or transport. Mixing of liquids is oftenachieved in stirred tanks. Mixing of wet and dryingredients can be carried out in kneading ma-chines for doughs or in stirred tanks for less vis-cous suspensions.

Bioprocessing Operations

For microbial and enzymatic transformations,a first requirement is the presence of the requiredingredients, including the desired microorgan-ism^) and/or enzyme(s), the required substratesand cofactors, and sufficient water. Several ofthe unit operations mentioned above will be in-volved to fulfill these requirements.

In order to allow the transformations to takeplace, incubation under optimum conditions forthe correct period of time will be needed. In or-der to safeguard the constancy of these incuba-tion conditions, unit operations such as mixing,heating, cooling, and transport are needed to en-sure even distribution of mass and heat, compen-sate for heat losses or generated heat, and com-pensate for deficiencies or overproduction ofmass by additions or removal.

Thermal Processing Operations

Heating and cooling are both characterized asthe transport of heat. Heat treatments and cool-ing are of extreme importance in food process-

ing and thus merit specific attention. The pri-mary objective of heating is to render food palat-able. It causes the gelatinization of starch, dena-turation of protein, and softening of toughtissues, and transforms a number of flavors.

Heat treatments consist of a warming-upphase, a period of holding time, and a cooling-down phase. Temperatures exceeding 70 0Ccause enzyme inactivation and kill vegetativemicrobial cells. The combination of temperatureand time determines the lethal effect of heattreatments. In principle, the term pasteurizationcorresponds to mild heat that kills heat-sensitivevegetative cells of bacteria, yeasts, and molds.Sterilization signifies killing all living cells, in-cluding more heat-resistant spores of bacteriaand molds. In food processing practice, steriliza-tion is not always required to ensure long-termshelf life of foods. In this respect, the term com-mercial sterility indicates a situation wheresome heat-resistant spores may have survivedthe heat treatment, but the composition of thefood prevents their revival during storage.

In view of bioprocessing, the timing of heattreatments is of crucial importance. Fermenta-tion of ingredients without prior heat treatment(e.g., raw cereals, raw meat, etc.) has by defini-tion a mixed character: Microorganisms and en-zymes that were present in the ingredient willtake part in the fermentation, sometimes as soleactors and sometimes as an accompaniment toadded starter microorganisms. At larger scaleproduction, fermentations must be better definedand controlled, and selected or optimized startercultures must be used. In order to ensure opti-mum functioning and to remove potentially haz-ardous microorganisms, the ingredients can bepasteurized or sterilized prior to inoculation andfermentation. In such case, it should be realizedthat some heat-sensitive growth factors such asenzymes and vitamins may have to be replenishedin order to enable effective microbial metabolism.

Cooling in food fermentation can be used as aprocessing tool. Prior to inoculation, heat-treated ingredients must be cooled to inoculationtemperature. During fermentation, the excessiveproduction of metabolic heat must be removedby cooling in order to prevent fermentation fail-

ures. In certain products such as yogurt, the fer-mentation is halted by cooling in order to pre-vent excessive acid development. Not all fer-mented foods have a long shelf life. Refrigerateddistribution and storage contribute to shelf lifeand hygienic safety.

Organization in Flow Diagram

A flow diagram is a pictorial scheme illustrat-ing the sequence in which ingredients and unitoperations are combined and providing data re-garding processing conditions such as tempera-ture, time, mass, pH, and so forth. Figure 1-7 il-lustrates this principle.

Appendix 1— A provides flow diagrams for se-lected food fermentations. For each major food

group, two commodities have been selected.These flow diagrams are not intended to providerecipes or exclusive methods of preparing therespective products. Their purpose is to providean insight into the process conditions and tim-ing, as well as relevant environmental antimicro-bial conditions that are of importance in foodsafety.

PROCESS CONDITIONS

Appendix 1— A aims to provide concrete pa-rameter values (or ranges) affecting microbialgrowth, metabolism, survival, and death as wellas enzymatic activity/stability, that are relevantto food safety. These data can be useful in

cereal grains

grading, cleaning

grinding

flour

mixing, kneading

dough

fermentation 2-3 h, 25-30 0C

baking 20-30 min, 220-250 0C

bread

cooling to ambient temperature

packaging, distribution

consumption

water, salt, fat

baker's yeast

Ingredients, (intermediate) products

Physical unit operations

Bioprocessing unit operations

Thermal unit operations

Figure 1-7 Principle of flow diagram (case of bread making)

HACCP approaches. Appendix 1— A providesflow diagrams and process conditions of rel-evance for the processing of starchy roots andtubers (cassava, Tables 1-A-l and l-A-2), ce-reals (barley, rye, and wheat, Tables l-A-3 andl-A-4), legumes (soybeans, Tables l-A-5 andl-A-6), vegetables (cabbage, Table l-A-7 andolives, Table l-A-8), fruits (grapes, Table 1-A-9 and palm sap, Table 1-A-l O), meat (pork,Tables 1-A-l 1 and l-A-12), fish (fresh water,Table l-A-13 and sea water, Table l-A-14),and milk (yogurt, Table 1-A-l 5 and cheese,Table 1-A-l 6). The icons explained in Figure1-7 are used throughout these flow diagrams toprovide a quick overview of the sequence of bio-and thermal unit operations. This is of specialimportance with regard to safety, as illustratedby the following cases.

REFERENCES

1. Beldman, G., Schols, H. A., Pitson, S. M., Searle-vanLeeuwen, M. J. F. & Voragen, A. G. J. (1997).Arabinans and arabinan degrading enzymes. In Ad-vances in Macromolecular Carbohydrate Research.Vol. 1, pp. 1-64. Edited by R. J. Sturgeon. London: JAIPress.

2. Belitz, H. -D. & Grosch, W. (1987). Food Chemistry.New York: Springer- Verlag.

3. Buckenhuskes, H. J. (1993). Selection criteria for lacticacid bacteria to be used as starter cultures for variousfood commodities. FEMS Microbiol Rev 12(1-3), 253-272.

4. Bylund, G. (1995). Dairy Processing Handbook. Lund,Sweden: Terra Pak Processing Systems AB.

5. Campbell-Platt, G. (1987). Fermented Foods of theWorld: A Dictionary and a Guide. Guildford, Surrey,UK: Butterworth Scientific.

6. Campbell-Platt, G. & Cook, P. E., eds. (1995). Fer-mented Meats. London: Blackie Academic & Profes-sional.

7. Carpita, N. C. & Gibeaut, D. M. (1993). Structural mod-els of primary cell walls in flowering plants: consistencyof molecular structure with the physical properties of thewalls during growth. The Plant Journal 3, 1-30.

8. Corroler, D., Mangin, I., Desmasures, N. & Gueguen,M. (1998). An ecological study of lactococci isolatedfrom raw milk in the Camembert cheese registered des-ignation of origin area. Appl Environ Microbiol 64(12),4729^735.

• Case 1: Is the food heated at all, some-where during the process? This may be ofcrucial importance in view of viruses (referto Chapter 8) and parasites (Chapter 9).

• Case 2: If heated, does this take place beforeor after fermentation? This will have seriousimplications in view of the activity of endog-enous enzymes, which may detoxify endog-enous toxic substances (Chapter 4), and inview of naturally occuring pathogens (Chap-ter 7), as well as the need for added safestarter cultures (Chapters 10 and 11).

• Case 3: Is the fermented product usuallycooked prior to consumption (e.g., tempeh)or is it eaten uncooked (e.g., most yogurts)?Needless to say, cooking prior to consump-tion will provide an additional "safety net"to the consumer.

9. Demuyakor, B. & Ohta, Y. (1993). Characteristics ofsingle and mixed culture fermentation of pito beer. J SdFoodAgric 62(4), 401^08.

10. Dittrich, H. H. (1995). Wine and brandy. In Biotechnol-ogy. Vol. 9, Enzymes, Biomass, Food and Feed, pp.463-504. Edited by G. Reed & T. W. Nagodawithana.Weinheim, Germany: VCH Verlagsgesellschaft.

11. Fleming, H. P., Kyung, K. H. & Breidt, F. (1995). Veg-etable fermentations. In Biotechnology. Vol. 9, En-zymes, Biomass, Food and Feed, pp. 629-661. Edited byH. J. Rehm, G. Reed, A. Punier & P. Stadler. Weinheim,Germany: VCH Verlagsgesellschaft.

12. Foschino, R., Terraneo, R., Mora, D. & Galli, A. (1999).Microbial characterization of sourdoughs for sweetbaked products. Ital J Food Sd 11(1), 19-28.

13. Fukushima, D. (1989). Industrialization of fermentedsoy sauce production centering around Japanese shoyu.In Industrialization of Indigenous Fermented Foods, pp.1-88. Edited by K. H. Steinkraus. New York: MarcelDekker.

14. Garrido Fernandez, A., Garcia, P. G. & Balbuena, M. B.(1995). Olive fermentations. In Biotechnology. Vol. 9,Enzymes, Biomass, Food and Feed, pp. 593—627. Editedby H. J. Rehm, G. Reed, A. Puhler & P. Stadler.Weinheim, Germany: VCH Verlagsgesellschaft.

15. Gervais, P., Molin, P., Grajek, W. & Bensoussan, M.(1988). Influence of the water activity of a solid sub-strate on the growth rate and sporogenesis of filamen-tous fungi. Biotechnol Bioengll, 457-463.

16. Gobbetti, M. (1998). The sourdough microflora: inter-actions of lactic acid bacteria and yeasts. Trends FoodSd Technol 9(7), 267-274.

17. Gripon, J. C. (1993). Mould-ripened cheeses. In Cheese:Chemistry, Physics and Microbiology. Vol. 2, MajorCheese Groups, pp. 1 1 1-136. Edited by P. F. Fox. Lon-don: Chapman & Hall.

18. Ko, S. D. & Hesseltine, C. W. (1979). Tempe and re-lated foods. In Economic Microbiology. Vol. 4, Micro-bialBiomass, pp. 1 15-140. Edited by A. H. Rose. Lon-don: Academic Press.

19. Kosikowski, F. V. (1982). Cheese and Fermented MilkFoods, 2nd edn. Ann Arbor, MI: Edwards Brothers.

20. Leistner, L. (1995). Stable and safe fermented sausagesworld- wide. In Fermented Meats, pp. 160—175. Editedby G. Campbell-Platt & P. E. Cook. London: BlackieAcademic & Professional.

21. Leistner, L. & Dresel, J. (1986). Die ChinesischeRohwurst - eine andere Technologic. MitteilungsblattBundesanstalt Fleischforschung 92, 6919-6926.

22. Leniger, H. A. & Beverloo, W. A. (1975). Food ProcessEngineering. Dordrecht, The Netherlands: D. Reidell.

23. Mittal, G. S. (1992). Food Biotechnology Techniquesand Applications. Lancaster, PA: Technomic.

24. Nabais, R. C., Sa- Correia, L, Viegas, C. A. & Novais, J.M. (1988). Influence of calcium ion on ethanol toleranceof Saccharomyces bayanus and alcoholic fermentationby yeasts. Appl Environ Microbiol 54(10), 2439-2446.

25. Nout, M. J. R. (1992). Ecological aspects of mixed-cul-ture food fermentations. In The Fungal Community: ItsOrganization and Role in the Ecosystem, 2nd edn, pp.817-851. Edited by G. C. Carroll & D. T. Wicklow.New York: Marcel Dekker.

26. Nout, M. J. R. (2000). Tempe manufacture: traditionaland innovative aspects. In Tempe. Edited by U.Baumann & B. Bisping. Weinheim, Germany: Wiley-VCH.

27. Nout, M. J. R., Martoyuwono, T. D., Bonne, P. C. J. &Odamtten, G. T. (1992). Hibiscus leaves for the manu-facture of usar, a traditional inoculum for tempe. .7 SdFoodAgric 58(3), 339-346.

28. Nout, M. J. R. & Rombouts, F. M. (1990). Recent devel-opments in tempe research. J Appl Bacteriol 69(5), 609-633.

29. Nout, M. J. R. & Rombouts, F. M. (2000). Fermentedand acidified plant foods. In The Microbiological Safetyand Quality of Food, pp. 685-737. Edited by B. M.Lund, T. C. Baird-Parker & G. W. Gould. Gaithersburg,MD: Aspen Publishers.

30. Nout, M. J. R. & Sarkar, P. K. (1999). Lactic acid foodfermentation in tropical climates. Antonie vanLeeuwenhoekl6, 395-401.

31. Patil, S. G. & Patil, B. G. (1989). Chitin supplementspeeds up the ethanol production in cane molasses fer-mentation. Enzyme Microb Technol 11, 38-43.

32. Phithakpol, B. (1993). Fish fermentation technology inThailand. In Fish Fermentation Technology, pp. 155-166. Edited by Ch. -H. Lee, K. H. Stemkraus & P. J. A.Reilly. Tokyo: United Nations University Press.

33. Rehm, H. J., Reed, G., Puhler, A. & Stadler, P. (1995).Biotechnology. Vol. 9, Enzymes, Biomass, Food andFeed. Weinheim, Germany: VCH Verlagsgesellschaft.

34. Rose, A. H. (1987). Responses to the chemical environ-ment. In The Yeasts. Vol. 2, Yeasts and the Environ-ment, 2nd edn, pp. 5^0. Edited by A. H. Rose & J. S.Harrison. London: Academic Press.

35. Russell, I. & Stewart, G. G. (1995). Brewing. In Biotech-nology. Vol. 9, Enzymes, Biomass, Food and Feed, pp.419-462. Edited by G. Reed & T. W. Nagodawithana.Weinheim, Germany: VCH Verlagsgesellschaft.

36. Samelis, J., Metaxopoulos, J., Vlassi, M. & Pappa, A.(1998). Stability and safety of traditional Greeksalami — a microbiological ecology study. Int J FoodMicrobiol 44(1-2), 69-82.

37. Seaby, D. A., McCracken, A. R. & Blakeman, J. P.(1988). Experimental determination of requirements forthe growth of edible Rhizopus species for use in solidsubstrate fermentation systems. J Sd Food Agric 44(4),289-299.

38. Spicher, G. & Brummer, J. -M. (1995). Baked goods. InBiotechnology. Vol. 9, Enzymes, Biomass, Food andFeed, pp. 241-319. Edited by G. Reed & T. W. Nagoda-withana. Weinheim, Germany VCH Verlagsgesellschaft.

39. Steinkraus, K. H. (1989). Industrialization of Indig-enous Fermented Foods. New York: Marcel Dekker.

40. Steinkraus, K. H. (Ed.). (1995). Handbook of Indig-enous Fermented Foods, 2nd edn. New York: MarcelDekker.

41. Suomalainen, H. & Oura, E. (1971). Yeast nutrition andsolute uptake. In The Yeasts. Vol. 2, Physiology andBiochemistry of Yeasts, pp. 3-74. Edited by A. H. Rose& J. S. Harrison. London: Academic Press.

42. Tamime, A. Y. & Robinson, R. K. (1985). Yoghurt: Sci-ence and Technology. Oxford, England: Pergamon.

43. Uhlig, H. (1991). Enzyme Arbeiten fir Uns. TechnischeEnzyme und Ihre Anwendung. Mimchen, Wien: CarlHanser Verlag.

44. Voragen, A. G. J. & Van den Broek, L. A. M. (1991).Fruit juices. In Biotechnological Innovations in FoodProcessing. Biotol Series, pp. 187-210. Edited byBiotol Team. Oxford, England: Butterworth-Heinemann.

45. Wood, B. J. B. (Ed.). (1998). Microbiology of Fer-mented Foods, 2nd edn. London: Blackie Academic andProfessional.

APPENDIX 1-A

Flow Diagrams for SelectedFood Fermentations

Table 1-A-1 Food Group: Starchy Roots and Tubers Product Name: Gari (Fermented and Dried Cassava) Reference: 40

Other Conditions ofAntimicrobial Relevance

(Salt, pH, Preservatives, etc.)

from 65% moisture content toapprox 50% moisture content

organic acids (lactic, acetic)0.6-1 .2% w/w

initial pH 6.2final pH 4-4.5

dehydration to approx 8%moisture content

in polythene bags or hermetic tins

Thermal Data (Time at Temp)

12-96 hat 25-35 0C

15-20 min at 80-85-100 0C

Ingredients and Microorganisms

cassava (Manihot esculenta)

natural fermentation dominated byLactobacillus plantarum, Coryne-bacterium spp., Geotrichum candidum

Flow Diagram

cassava tuberswashpeel (remove cortex)grate to coarse pulpput in jute or

polypropylene bagpressurize by weight to

enable de-wateringferment

break lumpsgarify (toasting for

dehydration andstarch gelatinization)

sievepackage and storereconstitute with cold

water or milkconsume

Note: Factors contributing to shelf life: Dehydration to about 8% moisture content, combined with hermetic storage allows shelf life of several months.Other remarks: During this process, the grating facilitates the enzymatic degradation and detoxification of naturally occurring cyanide (mainly in the form of linamarin). HCN (ppm)

levels in cassava roots, peeled roots, grated peeled roots, fermented pulp, and finished gari were 306, 184, 104, 52, and 10, respectively.

Table 1-A-2 Food Group: Starchy Roots and Tubers Product Name: Tape Ketella (Fermented Cassava) Reference: 40



ethanol to approx 2% v/vlactic acid 0. 1-0.3%final pH approx 5.0


1 0-30 min at 80-1 OO 0C

2-3 d at 25-30 0C


cassava (Manihot esculenta)

ragi tape (inoculum on rice powder)containing Amylomyces rouxii,Endomyces fibuligera, Endomycesburtonii, lactic acid bacteria

Flow Diagram

cassava tuberswashpeel (remove cortex)cut to 5-1 0 cm piecessteamcoolinoculate

ferment

consume

Note: Factors contributing to shelf life: Tape ketella has no shelf life. It is consumed as a snack or used occasionally as an ingredient in baked goods (cakes).

Table 1-A-3 Food Group: Cereals Product Name: Lager Beer Reference: 35



generation of saccharolytic,proteolytlc, and other brewingenzymes; partial modification ofthe barley grain

hops have antimicrobial effects

continues


1-2 d at 10-20 0C with inter-mittent airing, until moisturecontent 45-50% w/w4-6 d at 15-20 0C

at 71 -92 0C to reducemoisture content to 4-5% w/w

infusion mashing (30 min-several h at 65 0C) ordecoction mashing (2-3 h atincreasing temperatures from35-10O0C)

1-3 hat 100 0C

to 2-5 0C


barley (Hordeum vulgare)

product: sweet wortbyproduct: spent grain

hops (Humulus lupulus)

product: wort

Flow Diagram

1. Malting:soaking

germination

kilning

2. Mashing:coarse millingmashing

filtration

3. Wort boiling:hops additionboilingfiltrationcoolingfiltration

Table 1-A-3 continued


(Salt, pH, Preservatives, etc.)Thermal Data (Time at Temp)

1-2 wks at 10-1 5 0C

1-4 wks at 2-6 0C

cooling to 0-2 0C

5 min at 60 0C holding temp


brewers' yeast (Saccharomyces cerevisiaeor S. uvarum) at approx 107 yeast cellsper ml wort

product: young ("green") beer

product: mature beer

Flow Diagram

4. Fermentation:pitching

primaryfermentationsedimentation

secondaryfermentation(lagering)filtration

5. Bottling

6. Pasteurizationconsume

Note: Factors contributing to shelf life: Lager beer has no shelf life unless it is kept anaerobic and refrigerated. Bottled lager beer derives its shelf life from pasteurization and theexclusion of air.

Table 1-A-4 Food Group: Cereals Product Name: Wheat Mixed Grain Sourdough Bread Reference: 38



pH 4.6-4.7

0.5-0.7% lactic acid in bakedproduct


15-24 hat 23-31 0C

1-2 hat 26 0C

5 min

30-60 min at 30-40 0C35-40 min; oven temperature200-250 0Cto ambient temperature


rye flour 15 kgseed sour 0.3 kgwater 12.01yeasts (Candida krusei, Pichia saitoi,Saccharomyces cerevisiae, Torulopsisholmii) and lactic acid bacteria (Lacto-bacillus sanfranciscensis, Lb. plantarum,Lb. fermentum)

seed sour 27 kgrye flour 15 kgwheat flour 70 kgbakers' yeast 2 kgsalt (NaCI) 2 kgwater 51 Isee above, but dominated by Saccharo-myces cerevisiae (added bakers' yeast)

Flow Diagram

mix ingredientsfor sourdough

fermentation ofsourdough

seed sourmixing andkneading ofbread doughingredients

fermentation ofbread doughdivide, make-upintermediateproofshape and tintin proofbake

coolconsume

Note: Factors contributing to shelf life: Packaging in paper or other bags permitting moisture migration.Other Remarks: Storage at 20 0C for several days, limited by staling.

Table 1-A-5 Food Group: Legumes Product Name: Tempe Kedele (Soybean Tempeh) Reference: 27



pH soaking water decreased from6.5 to 4.1

ph beans 5.0-6.5

limited perforation allowsmycelium growth with very littlesporulationinitial pH 5.0;final pH 6.5-7.5NHs is formed due to proteolysis


6-24 h at 20-30 0C

20-40 min at 80-1 OO 0C

24-48 h at 25-30 0C externaltemp; internal temperaturemay become approx 1 0 0Chigher

few min in oil of 180 0C, or5-10 min in water or sauce of10O0C


soybeans (Glycine max), lactic acidbacteria (Lactobacillus plantarum), yeasts(Saccharomyces dairensis),Enterobacteriaceae

Rhizopus microsporus, R. Oligosporus, R.oryzae, Mucor indicusplant leaves (banana) or polyethylenesheets or boxes, with perforation holes

Flow Diagram

soak wholesoybeans in water

dehullcookcoolinoculate withmold starterpackage

ferment

harvest freshtempehslice or dicefry or stew

consume

Note: Factors contributing to shelf life: Fresh tempeh can be stored refrigerated or frozen.

Table 1-A-6 Food Group: Legumes Product Name: Koikuchi-Shoyu (Soy Sauce) Reference: 13



NaCL approximately 18% (w/v)

ethanol 2-3% (v/v), lactic acid1-2% (w/w), final pH 4.7-4.8


cooking: 40-45 min at 130 0Cor equivalent; roasting: fewmin at 170-1 80 0C

2-5 d at 25-30 0C

6-8 months

70-80 0C


soybeans (cooked, whole)wheat (roasted, crushed)

koji starter containing Aspergillus sojae,Asp. oryzae

kojibrineTetragenococcus halophila,Zygosaccharomyces rouxii

Flow Diagram

mixing equal quantitiesof soybeans and wheat

inoculation

incubation to obtainmolded mass = kojimixing koji and brineto obtain moromi mashfermentation

pressing,residue used asanimal feed

obtain rawsoy saucepasteurizebottleconsume

Note: Factors contributing to shelf life: Combined effect of high salt concentration, mild acidity, and pasteurization.

Table 1-A-7 Food Group: Vegetables Product Name: Sauerkraut (Sour Fermented Cabbage) References: 1 1 , 29



at least 1% w/w lactic acid shouldbe formed

4. canned sauerkraut

pasteurize 3 min at 74-820Chotfill in cans

continues


3-6 wks at 18-22 0C

3. fresh packaged long-lifesauerkrautadd preservative (0.1% w/wNa-benzoate)fill in glass or plastic


Brassica oleracea

byproduct: inner corebyproduct: outer leaves

add 1 .0-3.5 (typically 2.25)% w/w NaCI

using boards and weights or bags filledwith waternatural fermentation by microbialsuccession: Leuconostoc mesenteroides,Lactobacillus brevis, dominated byLactobacillus plantarum. Lactobacillusbavaricus occasionally added as a starterculture

2. fresh packaged sauerkraut

allow secondary fermentationto complete sugar depletionfill in plastic pouches

Flow Diagram

white cabbageharvest, transportgradingcoringtrimmingshredding to 1-2 mmthicksalting (andhomogenous mixing)load into fermentationtanks or vatscover with sheetheading (place weighton cabbage)fermentation

Postfermentation Options1 . fresh sauerkraut inbulk

refrigerate and sellshelf life 1 8-30 months

consume

refrigerate and sellshelf life 8-1 2 months

consume

refrigerate and sellshelf life> 2 months at < 7 0C; > 3months under CCk modifiedatmosphereconsume

Table 1-A-7 continued

refrigerate and sellshelf life 1-2 wks

consume

Note: Factors contributing to shelf life: The combination of acidity and depletion of fermentable sugars. In addition, exclusion of air, preservatives, pasteurization, or sterilization, asshown above.

Other remarks: Sauerkraut can be eaten without any heat treatment (e.g., in sandwiches), but it is also popular in cooked dishes with potatoes and meats.

Table 1-A-8 Food Group: Vegetables Product Name: Treated Black Olives in Brine References: 14, 29



fruits contain 0.5-1 .0% organicacids (citric, oxalic, malic), as wellas 1-3% phenolic compoundswith antimicrobial effect

pH to 5.8-7.9

final NaCI concentration approx1 .5% w/w

Thermal Data (Time at Temp)Ingredients and Microorganisms

Olea europaea sativa of various stages ofmaturation (yellowish green to purple)

5-7% w/v NaCI

natural fermentation; no starter added;lactic acid bacteria (dominated byLactobacillus plantarum) and yeasts(Pichia membranifaciens, Pichia vinii)1-2% w/v NaOH

purging with air; add iron salts to stabilizeblack color

2.5-5.0% w/v NaCI

add 1 .5% w/v NaCI brine

Flow Diagram

fresh olives

wash

brine

ferment

lye treatment

air oxidation(blackening)wash, neutralize

storage in brine

sort, sizecan

sterilizeconsume

Note: Factors contributing to shelf life: The combination of heat treatment and moderate salt and acidity. The antimicrobial effect of phenolic compounds occurring in fresh olives is notrelevant for the shelf life of the finished product because the mentioned substances have been degraded during the lye treatment.

Other remarks: Canned olives are often consumed without prior heating.

Table 1-A-9 Food Group: Fruits Product Name: Red Grape Wine References: 5, 10, 25



fermentation until fermentable

sugars less than 0.1%. Finalethanol 1 1-17% v/v. Acidity:volatile 0.1-0.15% w/v as aceticacid; total acidity 0.5-0.7% w/v astartaric acid

conversion of malic acid into lacticacid (typically 1 .5-3.5 g/l lacticacid)


1-3 wks at 20-25 0C

few days at pH > 3.5 andtemp > 15 0C

3 months-2 years


Vitis vinifera

SO2 100-1 50 mg/lsometimes additional nitrogen sources

starter cultures are occasionally used;otherwise natural fermentation by succes-sion; wild yeasts (Kloeckera apiculata,Kloeckera apis, Torulopsis stellata,Candida stellata), followed by Saccharo-myces cerevisiae (dominating primaryfermentation)separate skins ("lees") from young winestarter cultures used: Oenococcus oeni

bulk tanks

Flow Diagram

grapesdestemmingcrushing, sulfiting

fermentation "on theskins"

rackingsecondary fermentation

clarificationagingblendingfiltrationbottlingconsume

Note: Factors contributing to shelf life: The combination of ethanol, moderate acidity, and exclusion of air. Residual sulfite may be present, but this is not an essential preservative.Absence of fermentable sugars ("dry wines") contributes to shelf life.

Table 1-A-10 Food Group: Fruit Product Name: Toddy (Palm Wine) Reference: 40



organic acids (lactic, acetic)0.2-0.4%pH 4.5-3.5ethano!4-10%


0-48 h at 25-35 0C

15minat80°C


sap of oilpalm (Elaeis guineensis) orcoconut palm (Cocos nucifera)yeasts (Saccharomyces cerevisiae,Candida spp.), lactic acid bacteria(Lactobacillus spp., Leuconostoc spp.),acetic acid bacteria, Zymomonas spp.,Micrococcus spp.

Flow Diagram

collection of palm sap

natural fermentation

filtrationbottle (optional)pasteurize (optional)cool (optional)consume

Note: Factors contributing to shelf life: Fresh palm wine has no shelf life except when it is kept refrigerated or bottled and pasteurized.

Table 1-A-11 Food Group: Meat Product Name: Salami (Italian-Style Raw Fermented Sausage) Reference: 6



acidity approx 2.5% w/w as lacticacid, various antimicrobialcompounds (e.g., HbO2), pH initial5.8-6.0 to pH final 4.9-6.0


15-9Od at 15-25 0C

several wks at 12-15 0C toreduce Aw to 0.67-0.92


pure frozen pork or a pork/beef mix, fat,salt (NaCI) 2.5-4.5% w/wNaNO21 25 ppmpepper, mace, cardamon, garlic totalling1%w/wnatural or artificial casings

Lactobacillusspp., Pediococcusacidilactici, Micrococcusspp.,Debaryomyces hansenii

Flow Diagram

medium chopping(cuttering)

filling into casings

ripening (fermentation)

dryingconsume

Note: Factors contributing to shelf life: The combination of salt, nitrite, acidity, pH, anaerobic conditions, and reduced water activity. The lower the A ,̂, the longer the shelf life. Thistype of sausage is not smoked, in contrast to many other raw fermented sausages.

Other remarks: If no frozen pork was used, it must be heated to 58 0C to kill Trichinella. Otherwise, no heat treatment takes place and the product is consumed without prior cooking.

Table 1-A-12 Food Group: Meat Product Name: Lup Cheong (Chinese raw sausage) References: 20, 21



final pH 5.9 and Aw 0.8taste preference requires lowlevels of acidity


approx 36 h at 48 0C and65% RH to Aw 0.93 d at 20 0C and 75% RH


pork meat and fat, sugar, salt

natural casing (intestine)

no starter; natural fermentation by lacticacid bacteria, micrococci, etc.

Flow Diagram

coarse chopping

filling into casings

predrying

fermentation

cookingconsume

Note: Factors contributing to shelf life: The combination of reduced water activity, salt, and moderate acidity.Other remarks: The product is heated prior to consumption, distinguishing it from the consumption of European-style raw sausage.

Table 1-A-13 Food Group: Fish Product Name: Plaa-Raa (Fish-Rice Paste) Reference: 32



salt11-24%w/w

acidity 0.7-1 .9% as lactic acid;final pH varies from 4.1-6.9


several days at ambienttemperature

several days to a year


gourami (Trichogasteaspp.), snake-headfish (Ophicephalus striatus), catfish(Clarias spp., Heterobagrus sp.), tawes,bards (Puntius spp.), butter catfish(Ompoksp.), tilapia (Tilapia spp.)NaCI

endogenous proteases, Tetragenococcushalophila, Staphylococcus epidermidis,Micrococcusspp., Bacillus spp.glutinous or normal paddy rice, roast untildark brown, coarse grinding

Tetragenococcus halophila and otherlactic acid bacteria, Staphylococcusepidermidis, Micrococcusspp., Bacillusspp.

Flow Diagram

small fresh waterfish (whole)

mix with salt

ferment

mix with roasted rice

pack in earthen orglass jarsferment

consume

Note: Factors contributing to shelf life: Combined effect of high salt concentration and hermetic storage.

Table 1-A-14 Food Group: Fish Product Name: Nuoc-Mam (Fish Sauce) Reference: 40



lactic acid approx 1 % w/w;final pH 5.7-6.0

salt content 24-28% w/v


few months to 1 year at20-30 0C


Decapterus, Engranlis, Dorosoma,Clupeodes, Stolephorus spp.fish :salt 1:3 to 1:1. 5

endogenous proteases (halotolerantbacteria: Paracoccus halodenitrHicans,Aerococcus haloviridans are present; theirfunction is not clear)

Flow Diagram

small sea fish, whole

mix with salt

transfer to large vesselcover with layer of salt

pressurize by weightfermentation

drain liquid to obtainfirst-quality Nuoc-Mam

wash residue withboiling sea water inseveral stages toobtain second andsubsequent qualitiesbottlingconsume

Note: Factors contributing to shelf life: The combination of high salt content, organic acids, and hermetic bottling ensure a shelf life of several months.

Table 1-A-15 Food Group: MIIk Product Name: Stirred Yogurt Reference: 42



organic acids (lactic, acetic) toapprox 0.8-1% w/w as lactic acid;final pH 4.1-4.6


55 0C at 20 MPa5 min at 85 0C

16-20 hat 32 0C


cow milk

0.025% w/w starter culture containingequal numbers of Streptococcus salivariusspp. thermophilus and Lactobacillusdelbrueckiispp. bulgaricus

Flow Diagram

milkstandardizehomogenizepasteurize (highpasteurization)cool to 30-32 0Cinoculate

fermentstircool to 6 0Caseptic packagingconsume

Note: Factors contributing to shelf life: The combination of acidity and other antimicrobial compounds produced by lactic acid bacteria, anaerobic conditions in hermetically sealed jar,low storage temperature, and the presence of live (competing) microorganisms.

Other remarks: The product is consumed without prior heating.

Table 1-A-16 Food Group: Milk Product Name: Camembert Surface-Ripened Cheese References: 4, 17, 19



until 0.2% titratable acidity and pH4.5-4.6

whey pH 5.5

pH cheese increases to 6-7


20 sec at 70-72 0C

15-60minat34°C

2-3 h at 34 0C

24 hat 18 0C

5-6 h2-3 d at 14 0C1-1 .5 hat 14 0C8 d at 13-16 0C at 95% RH3 wks at 1 1 0C at 85% RH


cow milk 48% fat

mesophilic lactic acid bacteria cheesestarter Lactococcus lactis spp. cremoris,Lc. lactis spp. lactis, Leuconostocmesenteroides spp. cremoris, Leuc. lactis

rennet 0.01 % v/v

spore suspension Penicillium camemberti

2-3 times by rubbing crystalline saltor in 25% brine

aluminium foil, polyethylene film, woodenor cardboard boxes

Flow Diagram

pasteurization

coolinginoculation

stir

renetting

cut, drain wheyspray-inoculate

superficial dryingsalting

ripening first phaseripening second phasepackaging anddistributionconsume

Note: Factors contributing to shelf life: Refrigerated storage at 2-4 0C.

CHAPTER 2

Why Fermented Foods Can Be SafeMartin R. Adams

INTRODUCTION

Over time, the human diet has evolved to ex-clude materials that are frankly hazardous andfor which there is no simple procedure to renderthem harmless. As a result, our food supply isgenerally safe, although it can never be entirelydevoid of risk. There are a variety of hazards thatcan be associated with food materials. These canbe broadly classified into three categories: in-trinsic, extrinsic, and processing/biogenic (Ex-hibit 2-1).

Some of these hazards can be severe, as in thecase of botulinum toxin or bongkrekic acid poi-soning, but they can be controlled, and the riskthey pose can be reduced to acceptable levels byeffective and hygienic processing and prepara-tion. Safety considerations should always beparamount in modern food processing and arealso an implicit aspect of traditional methods ofprocessing and preserving food. This is particu-larly true of fermented foods, which take poten-tially hazardous raw materials such as raw meatand milk and transform them into acceptableproducts with better keeping qualities and re-duced risk. The bulk of this book addresses par-ticular groups of hazards, their significance infermented foods, and ways in which they arecontrolled. This chapter addresses the mecha-nisms by which fermentation can improve foodsafety in more general terms.

PHYSICAL PROCESSING

In Chapter 1, the various physical unit opera-tions associated with fermented food productionwere described. These operations are common toa number of food production processes and,wherever they occur, they can have some impacton product safety.

Transport

The manner in which raw materials are trans-ported can affect their quality. Physical damageduring transport can breach natural antimicro-bial barriers, allowing microorganisms access tonutrient-rich underlying tissues, where they cangrow rapidly. Excessive delays during transportcan also increase time for microbial growth andfor natural processes of senescence to occur.

Grading and Sorting

Inspection and sorting to remove damaged orobviously infected raw material is a basic pro-tective barrier that can be useful in reducing risk.It can be used to exclude obviously infectedmeat, damaged or moldy fruit, or mold-infectedgrains. For example, mechanical or hand sortingof infected nuts has been shown to reduce over-all aflatoxin levels considerably; reduced afla-toxin levels were also evident after hand sorting

Exhibit 2-1 A Classification of Food Hazards

Intrinsic — Properties of the food itself,natural toxins (e.g., cyanogenic glycosides,glycoalkaloids)

Extrinsic — Contaminants posing a directhazard (e.g., infectious pathogens — Salmo-nella, Shigella, viruses, protozoa; toxicchemical contaminants — pesticide residues,heavy metals)

Processing/Biogenic — Hazards produced as aresult of processing and/or generated bymicrobial contaminants (e.g., nitrosamines,biogenic amines, bacterial toxins, mycotoxins)

of figs.82 This is, however, a relatively bluntweapon because, in many cases, there is no vi-sual indication to help distinguish foods thatpose a hazard from those that do not.

Cleaning and/or Washing

Microbial pathogens can be associated withthe surface of fruits and vegetables as a result ofcontamination in the field. These may be naturalsoil microorganisms such as the spore-formingpathogens, Bacillus cereus and Clostridum botu-linum, or Listeria monocytogenes, or they maybe enteric in origin, such as Salmonella, Escheri-chia coli, and Shigella, arising from contact withhuman or animal feces, sewage, or untreatedwater. Contaminants may also be introducedduring harvesting and postharvest handling.

Interior tissues are generally, but not invari-ably, sterile, and the majority of the surface-as-sociated microflora can be removed by washing.Typically, vigorous washing of plant products inclean water will reduce microbial levels by 10—100-fold, but does not ensure the completeelimination of risk or levels of risk reductioncomparable to processes such as pasteurizationor canning. The efficacy of washing can be im-proved slightly by incorporating antimicrobialssuch as chlorine in the wash water. For example,washing of lettuce leaves with water removed anaverage of 92.4% of the total microflora. This

was increased to 99.5% by including 100 mg I'1

of available free chlorine and adjusting the pH to4.5-5.0.3 Failure to remove more of the microf-lora was ascribed to the survival of bacteria in pro-tective hydrophobic pockets or folds in the leafsurface, an effect that will vary between com-modities as a result of differences in surface com-position and topology. Increased production ofready-to-eat vegetables has heightened interest inthe surface disinfection of fruits and vegetables,and the subject has been extensively reviewed.9

Washing of animal carcasses with hot solu-tions of lactic or acetic acid can reduce the totalmicrobial load by 2-3 log cycles, although uni-form treatment of all surfaces is difficult andSalmonella and E. coli show more marked resis-tance to such treatments.36

Physical Separation

Physical separation can include peeling ortrimming to remove external layers. Because theexternal layers are where most contaminationwill be found, removing them can substantiallyimprove overall microbiological quality pro-vided it is done hygienically, avoiding contami-nation of the freshly exposed surfaces under-neath. The interior tissues of plants are generallyconsidered to be sterile. Although this is not al-ways true, substantial reductions in count can beachieved by using only hygienically removedinterior tissues. With leafy vegetables such ascabbage, the bacterial load on inner leaves canbe 1000-fold lower than on the outer leaves andthe interior tissues of the plant.55

Chemical hazards can also be reduced by thephysical separation of plant components. The re-moval of green surface layers from potatoes canreduce their glycoalkaloid content, and detach-ing the 2-3 mm thick parenchymal layer of cas-sava that underlies the cortex will reduce theoverall concentration of the cyanogenic glyco-sides, linamarin and lotaustralin.51

Moisture Adjustment

Changes in the water activity (Aw) of a mate-rial can have a profound effect on microbial haz-

ards. Depending on the circumstances, a de-crease in Aw that is produced by drying or theaddition of solutes such as salt will reduce thepotential for microbial growth but also enhancemicrobial resistance to adverse conditions, thuspotentially enhancing the survival of pathogensthat may be present. Increasing the Aw has thepotential to allow growth of previously quies-cent organisms, perhaps allowing them to growto numbers that are sufficient to initiate an infec-tion once they are ingested or produce enoughtoxin in the product to cause illness. Under cer-tain circumstances, an increase in the Aw as a re-sult of soaking could also permit the onset of alactic fermentation, which will help control oreliminate microbial hazards.

Size Reduction

Comminution of a raw material will disruptcells, thereby releasing their contents. This, inturn, will increase the supply of nutrients to anymicroorganisms present, stimulating theirgrowth. This can be beneficial when the growthof those organisms whose dominance is neces-sary for a successful fermentation is encouraged,but in some circumstances, it could also stimulategrowth of any pathogens present. The precise out-come will depend on a number of factors such asrelative number of organisms present, their sub-strate affinities, and their physiological state.

The relative sensitivities of different compo-nents of the microflora to antimicrobial factorsproduced by the plant material may also be im-portant. Many plants have defense systems thatare activated by cellular disruption. These sys-tems release compartmentalized enzymes, al-lowing them to act on their substrates and pro-ducing compounds that are active againstmicroorganisms and predators such as insects.Examples of this are the release of isothio-cyanate by the myrosinase-glucosinolate systemin crucifers10 and allicin produced during thecrushing of garlic.

The reduction of chemical hazards is alsosometimes assisted by tissue disruption. Thecyanogenic glycosides in cassava are hydro-lyzed to produce a cyanohydrin by the endog-

enous enzyme, linamarase, when cellular dam-age occurs.18 Microorganisms appear to play norole in cyanogen reduction in the traditional fer-mented cassava product, gan,89 although in pro-cesses where whole roots or large pieces are fer-mented, microbial activity does assist theendogenous process by softening the tissues.91

Mixing

Mixing will serve to distribute microorgan-isms throughout a mass of material. Very often,this distribution will accelerate the desirable fer-mentation and thereby improve safety, althoughit also has the potential to transfer pathogens tomicroenvironments where they are protectedfrom antimicrobial factors that prevail else-where in the material.

MICROBIAL ACTIVITY

As already noted, the production of fermentedfoods shares many unit operations with othertypes of food processing, all of which can con-tribute to product safety. The unique feature offood fermentations, however, is the central rolethat microbial activity plays in the overall pro-cess, contributing a number of desirable proper-ties such as improved product shelf life, in-creased safety, and improved flavor or texture.In developed countries today, the availability ofmodern food preservation techniques, such as anefficient cold chain, have diminished the signifi-cance of fermentation as a food preservationtechnology, although it remains of major impor-tance in developing countries.

Improvements in food safety arising from mi-crobial activity during fermentation are largelydue to lactic acid bacteria (LAB), which are agroup of organisms that predominate in the ma-jority of fermented foods. Their growth and me-tabolism inhibit the normal spoilage flora of thefood material and any bacterial pathogens that itmay contain. This inhibition can act in twoways: It can slow or arrest growth of the organ-ism, or it can inactivate or kill the organism.Both procedures can result in an improvement insafety. With toxigenic pathogens, the inhibition

of growth can effectively ensure safety, assum-ing that initial numbers are below those neces-sary to produce levels of toxin that can cause ill-ness. With infectious bacterial pathogens,slowing or preventing growth may be insuffi-cient to guarantee safety because the infectiousdose of some pathogens can be very low. Inacti-vation or killing of bacteria does occur, but evenquite high levels of inactivation still may not besufficient to eliminate risk entirely. This will de-pend on factors such as the pathogen concerned,its initial numbers, and its physiological state.

The LAB

The LAB are a group of gram-positive, non-sporeforming, fermentative anaerobes that are of-ten aerotolerant (Table 2-1). They produce mostof their cellular energy as a result of the fermenta-tion of sugars. In the case of hexoses, this can pro-ceed by one of two pathways, providing a usefuldiagnostic feature as well as playing a critical rolein their antimicrobial activity. Homofermentersferment hexoses by the Embden-Meyerhof path-way to produce almost exclusively lactic acid;heterofermenters produce less acid overall as amixture of lactic acid, acetic acid, ethanol, andcarbon dioxide using the 6-phosphogluconate/phosphoketolase pathway.7

The ability of LAB to inhibit other organismshas been a topic of extensive research over theyears, and the field has been reviewed periodi-cally.4'40'52'72'74 A variety of antimicrobial factors

produced by LAB have been identified (Exhibit2-2); these will be discussed individually. Theproduction of many of these compounds is lim-ited to a few restricted species or strains of LAB.Because these strains/species are not ubiquitousin lactic fermented foods, it is clear that they canonly play some contributory role to the overallpreservative effect. For so many different LABto appear in fermented foods, the principal pre-servative factor must be something that is com-mon to them all. This central and unvarying fea-ture is their use of fermentative pathways togenerate cellular energy and the fact that thisleads inevitably to the production of organic ac-ids, principally lactic acid, and a decrease in pH.Acidity levels in some fermentations can exceed100 mM, reducing the pH to less than 4.0 inweakly buffered systems.

Organic Acids and Reduced pH

Low pH and the presence of organic acids arethe two principal components of the inhibition ofmicroorganisms under the acid conditions pro-duced by fermentation. The three contributoryaspects of acid inhibition were described nearly50 years ago by Ingram et al.42

1. thepH2. the degree of dissociation of the acid3. the inherent toxicity of the acid anion

Pathogenic bacteria will grow over a pH rangeof 2-5 units but generally grow best at pH valuesaround neutrality, in the pH range of 6-7. As the

Table 2-1 Principal Genera of Lactic Acid Bacteria Associated with Food

Rod Coccus Homofermenter Heterofermenter

Lactobacillus + + +Lactococcus + +Enterococcus + +Carnobacterium + - +Leuconostoc + - +Weissella + - +Oenococcus + - +Pediococcus + + -Streptococcus + + -Tetragenococcus + + -

Exhibit 2-2 Antimicrobial Factors Associated withLactic Acid Bacteria

Low pHOrganic acidsBacteriocinsCarbon dioxideHydrogen peroxideDiacetylEthanolReuterinNutrient depletion and crowding

pH decreases away from the optimum region,the growth rate declines, eventually reachingzero. At pH values below the minimum that willsupport growth, the microorganism is progres-sively inactivated. The rate of inactivation istemperature dependent — the higher the tempera-ture, the faster the rate. This relationship appliesfor temperatures from chill to lethal, and hasbeen noted in both food and model systems for anumber of pathogens such as Salmonella,E. coll, Listeria, and Yersinia enterocoliti-c<3,33'53'76'88 and also at low Aw.30

If microorganisms allowed their internal pHto equilibrate to the same value as their environ-mental pH, they would only be able to functionover a very limited pH range close to neutrality.In fact, for most pathogens, growth only ceasesonce the pH has dropped to less than pH 4.5.This is because bacteria have the ability to main-tain their internal pH higher than that of theiracidic environment. Both the cell membrane,which has a low permeability to hydrogen ions,and the intrinsic buffering capacity of the cyto-plasm help to maintain the cytoplasmic pH, butthe cell also possesses a number of activemechanisms for maintaining pH homeostasis.These include the removal of protons in ex-change for the uptake of potassium ions, and theuptake and decarboxylation of amino acids toproduce neutralizing amines.11'12'35

Although weak organic acids such as thoseproduced by fermentation are less effective atdecreasing the extracellular pH, they are moreeffective at inhibiting bacteria than strong acids

such as hydrochloric.15'20'64'83 For example, re-cent data have shown that for 19 strains ofenterohemorrhagic E. coli, the minimum growthpH was 4.25 when hydrochloric acid was used asthe acidulant and 5.5 when acetic acid was used.59

This observation is linked to two important prop-erties of acids such as acetic and lactic acid.

1. They are weak carboxylic acids that onlypartially dissociate in aqueous solution.

2. In their undissociated form, they carry nonet charge and have appreciable lipid solu-bility, which allows them to diffuse freelythrough the cell's plasma membrane down aconcentration gradient into the cytoplasm.

The fact that the undissociated species is themore active antimicrobial form of these acids isillustrated in Figure 2-1, where the degree of in-hibition of E. coll is clearly linked with the de-gree of dissociation of the acid present. In a fer-mented food, the low pH will increase theproportion of the undissociated form present.When the undissociated acid passes through thecell's cytoplasmic membrane into the higher pHof the cytoplasm, it will dissociate, therebyacidifying the cytoplasm and releasing the acidanion. The increased leakage of protons into thecytoplasm will place a metabolic burden on thecells, which will divert resources away fromgrowth-related functions, thus slowing growth.The cell will also accumulate the acid anion,which can disrupt cellular processes.79 As theextracellular pH decreases and the total concen-tration of acid increases, so the burden will in-crease until growth is no longer possible.

The pKa of an acid is therefore an importantdeterminant of an acid's antimicrobial activitybecause it describes the proportion present in theundissociated form at any given pH. Differencesin the lipophilicity of the undissociated acid andthe intrinsic toxicity of the anion will account fordifferences in the observed toxicity of acids withsimilar pKa values. Acetic acid (pKa 4.76) is aweaker acid than lactic acid (pKa 3.86); thiscould account for the frequent observation that itis a more effective antimicrobial.12 It may alsoaccount for the apparent synergy in the antimi-crobial effect of mixtures of lactic and acetic

Total acid (mmolar)Figure 2-1 Effect of total acid concentration on the percentage acid dissociation (x) and the specific growth rateof Escherichia coli (a) with acetic acid and (b) with lactic acid.

acid, where lactic acid, the stronger acid, de-creases the pH, thereby increasing the propor-tion of acetic acid in the undissociated form andthus potentiating its antimicrobial effect.2'78

Heterofermentative LAB can produce mixtures

of lactic and acetic acid when they have alterna-tive electron acceptors such as oxygen or fruc-tose to help regenerate their nicotinamideadenine dinucleotide (NAD). Although hetero-fermenters produce less overall acidity than

Gro

wth

rat

e (u

)G

row

th r

ate

(JLI)

% d

isso

ciat

ed%

dis

soci

ated

homofermenters, their early dominance in sev-eral natural vegetable fermentations could bevery important in providing the rapid inhibitionof other organisms present and setting the fer-mentation on its subsequent course.

The antimicrobial effect will depend on theamount of acid produced, which will depend onthe numbers of LAB present. Numerous studieshave shown that bacterial pathogens do not sur-vive well when they are added to a prefermentedfood where the LAB have had the opportunity togrow to large numbers and the pH is alreadylow.60'61'70'71'81'87 However, when pathogens arepresent at the start of fermentation, inhibitionwill be delayed until the LAB have achieved anumerical dominance and produced sufficientacid to achieve an effect. In one instance, whenthe LAB (in this case, Lactococcus lactis) out-numbered the pathogen (E. coli) by more than 5log cycles, the pathogen was still able to growfor five hours, increasing in numbers by 2 logcycles during that time.95

Bacteriocins

Bacteriocins are polypeptide antimicrobialsthat are produced by bacteria and are bacteri-cidal to other, normally very closely related, or-ganisms. In recent years, a considerable researcheffort has been devoted to the identification andcharacterization of bacteriocins produced byLAB. The widespread consumption of LAB infoods without any adverse health effects is takenas an indication that they can generally be re-

garded as safe1 and therefore their bacteriocinsmight have potential as "natural" food preserva-tives. Tangible practical results from this workhave been slight, though it has led to the descrip-tion of numerous new bacteriocins.43'50'66

These new bacteriocins have been classifiedinto four main groups (Table 2-2). Under thisscheme, Class I bacteriocins are the !antibiotics,which include nisin, the only bacteriocin that iscurrently approved for use in foods.29 Nisin is asmall (3.4 kDa) heat-stable polypeptide contain-ing some unusual amino acids introduced bypost-translational modification of a ribosomallysynthesized precursor. It is distinguished frommany other bacteriocins by its relatively broadrange of activity, being inhibitory to most gram-positive bacteria. It has been used as a food pre-servative for 50 years and has, more recently,been given "generally regarded as safe" (GRAS)status in the United States.69 Nisin' s principaluse in food processing has been to inhibit theoutgrowth of spores, which show the greatestsensitivity to nisin. Gram-positive non-sporeforming pathogens such as Staphylococcusaureus and L. monocytogenes are inhibited, al-though S. aureus is one of the most nisin-resis-tant gram positives and L. monocytogenes hasbeen shown to acquire nisin resistance quitereadily.26'39'62

Early work suggested that nisin resistance inseveral Bacillus species was due to the produc-tion of a nisinase enzyme that specifically re-duced the C-terminal dehydroalanyl lysine to

Table 2-2 Classification of Bacteriocins Produced by Lactic Acid Bacteria

Class Subclass Description

I Lantibiotics (e.g., nisin); small, heat stable, containing lanthionineII Small (<10 kDa), heat stable, non-lanthionine-containing,

membrane-active peptidesNa Pediocin-like peptides with Y-G-N-G-V-X-C- near the amino terminus;

anti-listeriallib Two-peptide bacteriocinsMc Thiol-activated peptides

III Large (>30 kDa) heat labile proteinsIV Complex bacteriocins: protein with lipid and/or carbohydrate

alanyllysine.45-47 More recently, attention has fo-cused on the composition of the cytoplasmicmembrane where nisin exerts its effect by form-ing pores. Studies have seen differences in thephospholipid content of membranes in nisin-re-sistant and nisin-sensitive strains of L. mono-cytogenes6^ and changes in the fatty acid com-position leading to less fluid membranes inresistant strains.57 This appears not to be thecomplete explanation, however, as several stud-ies have implicated other features of the cell en-velope in resistance. For example, it has beenshown that the hydrophobicity of L. monocyto-genes cells correlates with their nisin sensitiv-ity,27'63 that nisin-resistant strains show differentsensitivities to cell wall-acting compounds,22

and that removal of the cell wall of resistantstrains abolishes resistance.27 Some clarificationof these findings and the observation that cellspossess limited nisin-binding sites26 has comefrom the work of Breukink et al11 They con-firmed an observation made 20 years previ-ously75 that nisin binds with high affinity toLipid II, a membrane-anchored cell wall precur-sor, and proposed that differences in nisin sensi-tivity may reflect different amounts or availabil-ity of Lipid II.17

Nisin is used commercially as a partially puri-fied concentrate but may be produced in situduring a fermentation. Its production duringcheese fermentations and the consequent inhibi-tion of other starter organisms led to its first iso-lation and description. Studies of the effect ofnisin-producing cultures on the safety of fer-mented foods have produced mixed results. Insitu nisin production has been shown to inhibitthe growth of L. monocytogenes in Camembertcheese,56 although levels of 150 IU g-1 producedby Lc. lactis in fermented rice made no discern-ible contribution to the inhibition of S. aureus,which was ascribed entirely to the lactic acidproduced.95 However, there have been severalreports of work employing LAB-producing bac-teriocins other than nisin to inhibit L.monocytogenes in fermented sausages, cottagecheese, smoked salmon, and bean sprouts.8'34'58'68

The value of nisin and other LAB bacteriocinsas aids to bacteriological food safety is limited

by their range of activity. Gram-negative bacte-ria, which constitute the majority of food-bornebacterial pathogens, are generally resistant to ni-sin due to its inability to penetrate the outermembrane. If, however, the outer membranepermeability is increased by treatment with ch-elators such as ethylenediaminetetraacetic acid(EDTA) or thermal injury, then they do displaysensitivity.14'23'85'86 Prospects for using this ap-proach to inhibit gram negatives in fermentedfoods are not promising because the pH dropduring fermentation does not appear to produceappreciable outer membrane injury and reducesthe ability of EDTA to chelate metal ions andpermeabilize the outer membrane.13

Carbon Dioxide

Heterofermentative LAB produce 1 mole(22.4 liters) of carbon dioxide for every mole ofhexose they ferment. This can help establishanaerobic conditions, thereby preventing thegrowth of obligate aerobes such as molds andslowing the growth of facultative organismssuch as the Enterobacteriaceae. Anaerobic con-ditions will also give the LAB a competitiveedge and encourage a successful fermentation.An increased partial pressure of carbon dioxidealso has its own specific antimicrobial activity.Molds and oxidative gram-negative bacteria aremost susceptible, whereas lactobacilli and someyeasts show high tolerance. There is consider-able scope for improving the understanding ofthe inhibitory mechanisms of carbon dioxideand the relative importance of factors such as theability of carbon dioxide to decrease intracellu-larpH, its inhibition of enzymatic reactions, andthe disruptive effects of its interaction with cellmembranes.

Hydrogen Peroxide

In the presence of oxygen, LAB produce hy-drogen peroxide through the activity of fla-voprotein oxidases. Because LAB are catalasenegative, the hydrogen peroxide can accumulatein the medium. Hydrogen peroxide is a strongoxidizing agent and exerts its antimicrobial ef-fect directly or through its degradation productssuch as superoxide, Or, and hydroxyl radicals.

LAB appear to be more resistant to hydrogen per-oxide than many other bacteria. A reported mini-mum inhibitory concentration for L lactis was125 mg I-1, whereas that for S.aureus was 5-6mg I-1.25'92 However, the amounts accumulated byLAB in culture are quite variable. Lactococcigrowing in milk and the meat starter organism,Pediococcus cerevisiae, both produce levels be-low 1 mg I'1 in the bulk environment.73'77 Increasedamounts of hydrogen peroxide can be produced byimproving aeration through shaking and at lowtemperatures when the oxygen solubility isgreater.21'25 It has been suggested that hydrogenperoxide production by LAB in refrigerated foodsis a significant antimicrobial factor.16'37 However,this seems an unlikely scenario in the productionof most fermented foods. It is possible that hydro-gen peroxide plays a significant role during theearly stages of fermentation while residual dis-solved oxygen is removed, or that microenviron-ments with higher levels may persist. In general,though, the contribution of hydrogen peroxide islikely to be relatively minor.

Diacetyl

Diacetyl (2,3-butanedione) is produced bystrains of several LAB and plays an importantrole in the sensory properties of many fermentedfoods, particularly dairy products. It can accu-mulate when the organism reaches the stationaryphase of growth or when there are alternativesources available from which to generate the keyintermediate pyruvate, such as citrate. Antibac-terial activity of diacetyl has been demonstratedagainst a number of bacteria, including Aero-monas hydrophila, Bacillus spp., Enterobacteraerogenes, E. coli, Mycobacterium tuberculosis,Pseudomonas spp., Salmonella, S. aureus, andY. enterococolitica.31'44'4*'65 Gram negatives gen-erally show greater sensitivity and the LAB aremost resistant.

However, diacetyl does have an extremelysharp odor, and the acceptable sensory thresholdin dairy products is 2-7 mg kg'1.54 The levels ofdiacetyl reported to produce appreciable micro-bial inhition are usually quite high (e.g., forgram-negative bacteria, approximately 200 mgkg'1), well in excess of this level. Some studies

have suggested that lower concentrations ofdiacetyl can be effective, particularly at lowertemperatures,6 but others have failed to see anyappreciable contribution of diacetyl at 10 mg kg-1

to the inhibition of E. coli in refrigerated dairyfoods.38 Overall, the contribution of diacetyl tobacterial inhibition in fermented foods appearsto be relatively minor.

Ethanol

Ethanol is a well-established antimicrobial. Itis the principal fermentation product of the yeastSaccharomyces cerevisiae and plays a centralrole in the microbiological safety of alcoholicbeverages, where it is often the most importantsingle hurdle in a series that can also includeheat treatment, low pH, anaerobic conditions,and elevated carbon dioxide levels. Yeast activ-ity also removes the mycotoxin patulin fromapple juice (see Chapter 5).

Heterofermentative LAB will also produceethanol, but in lower concentrations. The levelsproduced are inversely related to the amount ofacetic acid produced because these compoundsrepresent alternative fates of acetyl phosphate inthe heterofermentation pathway. In addition toany direct inhibitory effect, ethanol has alsobeen shown to augment the lethal effect of lowpH and lactic acid in E. coli 0 157, though athigher concentrations than would be likely tooccur in lactic fermentations.49 Heterofer-menters dominate during the initial phase of sev-eral natural fermentations and ethanol may, likeother products of heterofermentation, contributein these early stages to setting the fermentationon its desirable course.

Reuterin and Other Low MolecularWeight Compounds

A number of low molecular weight compoundswith antimicrobial activity have been isolated andidentified from culture filtrates of LAB. The moststudied to date has been reuterin, p-hydroxy-propanal. This is present in stationary phase cul-tures of strains of the heterofermenter, Lactobacil-lus reuteri, when grown anaerobically on amixture of glucose and glycerol or glyceralde-hyde. Reuterin exists in three forms — the aide-

hyde, its hydrate, and a cyclic dimer — and hasvery broad spectrum antimicrobial activity againstbacteria, fungi, protozoa, and viruses. The relativeactivity of the different forms is not known, al-though they are thought to act by inhibiting sulfhy-dryl-containing enzymes.24

The addition of reuterin to foods such as mincedbeef, milk, and cottage cheese has been shown tocontrol coliform growth and to inactivate L.monocytogenes and.fi". coll O157.24'32 However, itsrole in fermented foods is uncertain. One studywhere L reuteri and glycerol were added to her-ring fillets to stimulate reuterin production in situshowed some potential to reduce growth of thegram-negative flora present. In other cases wherethe incorporation of L. reuteri in foods has beendescribed, in a fermented milk and an Emmental-type cheese, its presence has been for its potentialprobiotic effect rather than for reuterin productionin the food.28'84

Pyroglutamic acid, 2-pyrrolidone-5-carboxy-lic acid, is a natural constituent of some plantfoods and fermented products such as soy sauceand can be produced by the thermal dehydrationof glutamic acid.5'94 It has been shown to be pro-duced by a range of LAB, including strains of L.easel, L. helveticus, Streptococcus bovis, andunspecified pediococci.19'41'93 Pyroglutamic acidis active primarily against gram-negative bacte-ria such as Pseudomonas and Enterobacter. Itsactivity against gram-negative pathogens hasnot been reported, although it is inactive againstthe gram-positive pathogens L. monocytogenesand S. aureus.41

A number of other low molecular weight com-pounds with antimicrobial properties have been

identified in culture filtrates from L plantarumcultures. These include mevalonolactone, methyl-hydantoin, and benzoic acid.67 They have beenshown to act synergistically with lactic acidagainst Pantoea (Enterobacter) agglomerans, buttheir significance in food fermentations remains tobe established.

Nutrient Depletion and Overcrowding

The absence of inhibitory agents is a neces-sary but insufficient condition for microbialgrowth. The microbial cell must also have suffi-cient space and nutrients available to it. Foodswill rarely offer microorganisms a uniform me-dium, and numerous microenvironments mayexist where conditions differ substantially fromthe bulk properties of the material. This is a criti-cal factor in the microbiological stability of but-ter and similar spreads, where microorganismsare confined in aqueous droplets dispersed in thecontinuous fat phase. The presence of large num-bers of LAB in a fermented food will deplete it ofreadily available nutrients and occupy space thatmight have been available to other less desirableorganisms. This is supported by the recent obser-vation that filtration of a fermented broth to re-move cells of Lc. lactis decreased the subsequentinhibition of E. coli.13 It is an area that is only justbeginning to be explored.

OVERALL SIGNIFICANCE OFDIFFERENT ANTIMICROBIALFACTORS

To describe the combined effect of a numberof antimicrobial factors, the analogy of a hurdle

Table 2-3 Shelf Life and Composition of Fermented Fish Products from Thailand80

Product Salt(%) pH Shelf Life

Pla-som 2.3-5.9 4.0-4.6 3 weeksSom-fug 2.5-5.8 4.1-5.0 2 weeksPla-chom 3.8-4.8 5.0-6.1 2 weeksPla-chao 4.4-9.5 4.0-5.3 1-2 yearsPla-paeng-daeng 4.5-9.2 3.9-5.2 6-12 monthsPla-ra 7.8-17.9 4.7-6.2 1-3 years

race has been used. Perhaps nowhere in foodpreservation is the concept of multiple barriersor hurdles more appropriate than in fermentedfoods. The LAB themselves can contribute anumber of different antimicrobials, and micro-bial activity is often combined with other inhibi-tory factors such as salting or partial drying.

It is possible to make some generalizations re-garding the relative importance of these variousfactors. Certainly the predominant microbialfactor is the production of organic acids and thereduction in pH, although the others described

REFERENCES

1. Adams, M. R. (1999). Safety of industrial lactic acidbacteria. J Biotechnol 63, 171-178.

2. Adams, M. R. & Hall, C. J. (1988). Growth inhibition offood-borne pathogens by lactic and acetic acids andtheir mixtures. IntJFood Sd Technol 23, 287-292.

3. Adams, M. R., Hartley, A. D. & Cox, L. J. (1989). Fac-tors affecting the efficacy of washing procedures used inthe production of prepared salads. Food Microbiol 6,69-77.

4. Adams, M. R. & Nicolaides, L. (1997). Review of thesensitivity of different foodborne pathogens to fermen-tation. Food Control 8, 227-239.

5. Airaudo, C. B., Gayte-Sorbier, A. & Armand, P. (1987).Stability of glutamine and pyroglutamic acid undermodel system conditions: influence of physical andtechnological factors. J Food Sd 52, 1750—1752.

6. Archer, M. H., Dillon, V. M., Campbell-Platt, G. &Owens, J. D. (1996). Effect of diacetyl on growth rate ofSalmonella typhimurium determined from detectiontimes measured in a micro-well plate photometer. FoodControl 7, 63-67.

7. Axelsson, L. (1998). Lactic acid bacteria: classificationand physiology. In Lactic Add Bacteria: Microbiologyand Functional Aspects, pp. 1-72. Edited by S.Salminen & A. von Wright. New York: Marcel Dekker.

8. Bennik, M. H. J., van Overbeek, W., Smid, E. J. &Gorris, L. G. M. (1999). Biopreservation in modified at-mosphere stored mungbean sprouts: the use of vegetableassociated bacteriocinogenic lactic acid bacteria to con-trol the growth of Listeria monocytogenes . Lett ApplMicrobiol 28, 226-232.

9. Beuchat, L. R. (1998). Surface decontamination of fruitsand vegetables eaten raw: a review. Food Safety Unit,World Health Organization, WHO/FSF/FOS/98.2 42pp.

10. Bones, A. M. & Rossiter J. T. (1996). The myrosinase-glucosinolate system, its organisation and biochemistry.Physiol Plantarum 97, 194-208.

here could all contribute to the aggregate effect,particularly to ensuring a successful early domi-nance of LAB. Reduction in Aw generally playsa much more significant role in the inhibition ofmicrobial growth compared to bacterial acidproduction. This is illustrated by data on theshelf life of fermented fish products in Thailandand their composition, where differences in saltcontent have greater impact on product shelf life(Table 2-3). Microbial activity is, however,probably more important and effective at reduc-ing numbers of microbial hazards.

1 1 . Booth, I. R. (1999). The regulation of intracellular pH inbacteria. In Bacterial Responses to pH, pp. 19-37. Ed-ited by D. J. Chadwick & G. Cardew. Chichester, En-gland: John Wiley & Sons.

12. Booth, I. R. & Kroll, R. G. (1989). The preservation offoods by low pH. In Mechanisms of Action of FoodPreservation Procedures, pp. 1 19-159. Edited by G. W.Gould. London: Elsevier Applied Science.

13. Boziaris, I. S. & Adams, M. R. (1999). Effect of chela-tors and nisin produced in situ on inhibition and inacti-vation of gram negatives. Int J Food Microbiol 53, 105-113.

14. Boziaris, I. S., Humpheson, L. & Adams, M. R. (1998),Effect of nisin on heat injury and inactivation of Salmo-nella enteritidis PT4. Int J Food Microbiol 43, 7-14.

15. Brackett, R. E. (1987). Effects of various acids ongrowth and survival of Yersinia enteroccolitica. J FoodProt 50, 598-601.

16. Brashears, M. M., Reilly, S. S. & Gilliland, S. E. (1998).Antagonistic action of cells of Lactobacillus lactis to-ward Escherichia coli O157:H7 on refrigerated rawchicken meat. J Food Prot 61, 166-170.

1 7. Breukink, E., Wiedemann, L, van Kraaij, C., Kuipers, O.P., Sahl, H. -G. & de Kruijff, B. (1999). Use of the cellwall precursor Lipid II by a pore-forming peptide antibi-otic. Science 286, 2361-2364.

18. de Bruijn, G. H. (1973). The cyanogenic character ofcassava (Manihot esculenta}. In Chronic Cassava Tox-icity, pp 43^8. Edited by B. Nestel & R. Maclntyre.Ottawa, Canada: International Development ResearchCentre.

19. Chen, G. & Russell, J. B. (1989). Transport of glutamineby Streptococcus bovis and conversion of glutamine topyroglutamic acid and ammonia. JBacteriol 171, 2981-2985.

20. Chung, K. C. & Goepfert, J. M. (1970). Growth of Sal-monella at low pH. J Food Sd 35, 326-328.

21. ^ollins, E. B. & Aramaki, K. (1980). Production of hy-drogen peroxide by Lactobacillus acidophilus. J DairySd 63, 353-357.

22. Crandall, A. D. & Montville, T. J. (1998). Nisin resis-tance in Listeria monocytogenes ATCC 700302 is acomplex phenotype. Appl Environ Microbiol 64, 231—237.

23. Cutter C. N. & Siragusa G. R. (1995). Population reduc-tion of gram-negative pathogens following treatmentswith nisin and chelators under various conditions. JFood Prot 58, 977-983.

24. Daeschel, M. A. & Penner, M. H. (1992). Hydrogen per-oxide, lactoperoxidase systems and reuterin. In FoodBiopreservatives ofMicrobial Origin, pp. 155-175. Ed-ited by B. Ray & M. Daeschel. Boca Raton, FL: CRCPress.

25. Dahiya, R. S. & Speck, M. L. (1968). Hydrogen perox-ide formation by lactobacilli and its effect on Staphylo-coccus aureus. J Dairy Sd 51, 1568-1573.

26. Davies, E. A. & Adams, M. R. (1994). Resistance ofListeria monocytogenes to the bacteriocin nisin. Int JFood Microbiol 21, 341-347.

27. Davies, E. A., Falahee, M. B. & Adams, M. R. (1996).Involvement of the cell envelope of Listeriamonocytogenes in the acquisition of nisin resistance. JAppl Bacteriol 8 1 , 1 39-146.

28. Delespaul, G., Leclerc, G., Lepeltier, M. & Chassagne,M. H. (1994). Cheesemaking process and the cheesesobtained. French Patent 2,696,620.

29. Delves-Broughton, J. (1990). Nisin and its use as a foodpreservative. Food Technol 142, 100-1 17.

30. Deng, Y., Ryu, J. H. & Beuchat, L. R. (1998). Influenceof temperature and pH on survival of Escherichia coliO157:H7 in dry foods and growth in reconstituted infantrice cereal. Int J Food Microbiol 45, 173-184.

31. Elliker, P. R. (1945). Effect of various bacteria ondiacetyl content and flavour of butter. J Dairy Sd 28, 93.

32. El-Ziney, M. G. & Debevere, J. M. (1998). The effect ofreuterin on Listeria monocytogenes and Escherichia coliO157:H7 in milk and cottage cheese. J Food Prot 61,1275-1280.

33. Entani, E., Asai, M., Tsujihata, S., Tsukamoto, Y. &Ohta, M. (1998). Antibacterial action of vinegar againstfoodborne pathogenic bacteria. J Food Prot 61, 953-959.

34. Foegeding, P. M., Thomas, A. B., Pilkington, D. H. &Klaenhammer, T. R. (1992). Enhanced control of List-eria monocytogenes by in szYw-produced pediocin duringdry fermented sausage production. Appl EnvironMicrobiol 58, 884-890.

35. Foster, J. W. & Moreno, M. (1999). Inducible acid toler-ance mechanisms in enteric bacteria. In Bacterial Re-sponses topH, pp. 55-74. Edited by D. J. Chadwick &G. Cardew. Chichester, England: John Wiley & Sons.

36. Gill, C. O. (1998). Microbiological contamination ofmeat during slaughter and butchering of cattle, sheepand pigs. In The Microbiology of Meat and Poultry, pp.118-157. Edited by A. Davies & R. Board. London:Blackie Academic and Professional.

37. Gilliland, S. E. & Ewell, H. R. (1983). Influence of com-binations of Lactobacillus lactis and potassium sorbateon growth of psychrotrophs in raw milk. J Dairy Sd 66,974-980.

38. Guraya, R., Frank, J. F. & Hassan, A. N. (1998). Effec-tiveness of salt, pH and diacetyl as inhibitors for Es-cherichia coli O157:H7 in dairy foods stored at refrig-eration temperatures. J Food Prot 61, 1098-1 102.

39. Harris, L. J., Fleming, H. P. & Klaenhammer, T. R.(1991). Sensitivity and resistance of L. monocytogenesATCC 19115, Scott A and UAL 500 to nisin. J FoodProt 54, 836-840.

40. Holzapfel, W. H., Geisen, R. & Schillinger, U. (1995).Biological preservation of foods with reference to pro-tective cultures, bacteriocins and food-grade enzymes.Int J Food Microbiol 24, 343-362.

41. Huttunen, E., Noro, K. & Zang, Y. (1995). Purificationand identification of antimicrobial substances producedby two Lactobacillus casei strains. Int Dairy J 5, 503-513.

42. Ingram, M., Ottaway, F. J. H. & Coppock, J. B. M.(1956). The preservative action of acid substances infood. Chem and Ind 42, 1154-1165.

43. Jack, R. W., Tagg, J. R. & Ray, B. (1995). Bacteriocinsof gram positive bacteria. Microbiol Rev 59, 171-200.

44. Jalander, Y. W. (1936). Diacetyl als Tuberkelbazillentotender Bestandteil des finnischen Holzteeres. Naunyn-Schmiedebergs Arch Experimental Pathol Pharmakol180, 628-630.

45. Jarvis, B. (1967). Resistance to nisin and production ofnisin-inactivating enzymes by several Bacillus species.J Gen Microbiol 47, 33-48.

46. Jarvis, B. (1970). Enzymic reduction of the C-terminaldehydroalanyl-lysine sequence in nisin. Proc Biochem£?cll9,56P.

47. Jarvis, B. & Farr, J. (1971). Partial purification, specific-ity and mechanism of action of the nisin-inactivatingenzymes from Bacillus cereus. Biochim Biophys Acta227, 232-240.

48. Jay, J. M. (1982). Antimicrobial properties of diacetyl.Appl Environ Microbiol 44, 525-532.

49. Jordan, S. L., Glover, J., Malcolm, L., Thomson-Carter,F. M., Booth, I. R. & Park, S. F. (1999). Augmentationof killing of Escherichia coli 0 157 by combinations oflactate, ethanol, and low pH conditions. Appl EnvironMicrobiol 65, 1308-1311.

50. Klaenhammer, T. R. (1993). Genetics of bacteriocinsproduced by lactic acid bacteria. FEMS Microbiol Rev12, 38-85.

51. Kojima, M., Iwatsuki, N., Data, E. S., Villegas, C. D. V.& Uritani, I. (1983). Changes of cyanide content andlinamarase activity in wounded cassava roots. PlantPhysioll2, 186-189.

52. Lindgren, S. E. & Dobrogosz, W. J. (1990). Antagonis-tic activities of lactic acid bacteria in food and feed fer-mentations. FEMS Microbiol Rev 87, 149-164.

53. Little, C. L., Adams, M. R., Anderson, W. A. & Cole, M.B. (1994). Application of a log-logistic model to describethe survival of Yersinia enterocolitica at sub-optimal pHand temperature. IntJFood Microbiol 22, 63-7 1 .

54. Liicke, F. -K. & Earnshaw, R. G. (1991). Starter cul-tures. In Food Preservatives., pp. 215-234. Edited by N.Russell & G. W. Gould. Glasgow, Scotland: BlackieAcademic and Professional.

55. Lund, B. M. (1992). Ecosystems in vegetable foods. JAppl Bacterial Symp Supp 73, 1 15S-126S.

56. Maisnier-Patin, S., Deschamps, N., Tatini, S. R. & Rich-ard, J. (1992). Inhibition of Listeria monocytogenes inCamembert cheese made with a nisin-producing starter.ZfljY 72, 249-263.

57. Mazzotta, A. S. & Montville, T. J. (1997). Nisin induceschanges in membrane fatty acid composition of Listeriamonocytogenes nisin-resistant strains at 10 0C and 300C. J Appl Microbiol 82, 32-38.

58. McAuliffe, O., Hill, C. & Ross, R. P. (1999). Inhibitionof Listeria monocytogenes in cottage cheese manufac-tured with a lacticin 3147-producing starter culture. JAppl Microbiol 86, 251-256.

59. McKellar, R. C. & Knight, K. P. (1999). Growth andsurvival of various strains of enterohemorrhagic Es-cherichia coli in hydrochloric and acetic acid. J FoodProt 62, 1466-1469.

60. Mensah, P., Tomkins, A. M., Drasar, B. S. & Harrison,T. J. (1990). Fermentation of cereals for reduction ofbacterial contamination of weaning foods. Lancet 336,140-143.

61. Mensah, P., Tomkins, A. M., Drasar, B. S. & Harrison,T. J. (1991). Antimicrobial effect of fermented Ghana-ian maize dough. J Appl Bacteriol 70, 203-210.

62. Ming, X. & Daeschel, M. A. (1993). Nisin resistance offoodborne bacteria and specific resistance responses ofListeria monocytogenes Scott A. J Food Prot 56, 944-948.

63. Ming, X. & Daeschel, M. A (1995). Correlation of cellu-lar phospholipid content with nisin resistance of Listeriamonocytogenes Scott A. J Food Prot 58, 416^20.

64. Minor, T. E. & Marth, P. H. (1972). Loss of viability ofStaphylococcus aureus in acidified media: inactivationby several acids, mixtures of acids and salts of acids. JMilk Food Technon5, 191-196.

65. Motlagh, A. M., Johnson, M. C. & Ray, B. (1991). Vi-ability loss of foodborne pathogens by starter culturemetabolites. J Food Prot 54, 873-878.

66. Nes, I. F., Diep, D.B., Havarstein, L. S., Brurberg, M.B., Eijsink, V. & HoIo, H. (1996). Biosynthesis of bacte-riocins in lactic acid bacteria. Antonie van Leeuwenhoek70,113-128.

67. Niku-Paavola, M. -L., Laitila, A., Mattila-Sandholm, T.& Haikara, A. (1999). New types of antimicrobial com-pounds produced by Lactobacillus plantarum. J ApplMicrobiol 86, 29-35.

68. Nilsson, L., Gram, L. & Huss, H. H. (1999). Growthcontrol of Listeria monocytogenes on cold-smokedsalmon using a competitive lactic acid flora. J Food Prot62, 336-342.

69. Nisin preparation: affirmation of GRAS status. (1988).Federal Register 54, 1 1247-1 125 1 .

70. Nout, M. J. R., Rombouts, F. M. & Hautfast, G. J.(1989). Accelerated natural lactic fermentation of infantfood formulations. Food Nutr Bull 1 1 , 65-73.

71. Nout, M. J. R., Rombouts, F. M. & Havelaar, A. (1989).Effect of accelerated natural lactic acid fermentation ofinfant food ingredients on some pathogenic organisms.Int J Food Microbiol 8, 351-361.

72. Ouwehand, A. (1998). Antimicrobial components fromlactic acid bacteria. In Lactic Acid Bacteria: Microbiol-ogy and Functional Aspects, pp. 139-159. Edited by S.Salminen & A. Von Wright. New York: Marcel Dekker.

73. Raccach, M. & Baker, R. C. (1978). Formation of hy-drogen peroxide by meat starter cultures. J Food Prot.41,798-801.

74. Ray, B. & Daeschel, M., eds. (1992). Food Biopreserva-tives ofMicrobial Origin. Boca Raton, FL: CRC Press.

75. Reisinger, P., Seidel, H., Tschesche, H. & Hammes, W.(1980). The effect of nisin on murein synthesis. ArchMicrobiol 127, 187-193.

76. Roering, A. M., Luchansky, J. B., Ihnot, A. M., Ansay,S. E., Kaspar, C. W. & Ingham, S. C. (1999). Compara-tive survival of Salmonella typhimurium DT 104, List-eria monocytogenes and Escherichia coli O157:H7 inpreservative-free apple cider and simulated gastric fluid.Int J Food Microbiol 46, 263-269.

77. Ross, G. D. (1981). The inhibition of growth of spoilagemicroorganisms in milk by Streptococcus lactis subsp.diacetylactis, Leuconostoc cremoris and L.dextranicum. AustJ Dairy Technol 36, 147-149.

78. Rubin, H. E. (1973). Toxicological model for a two-acidsystem. Appl Environ Microbiol 36, 623-624.

79. Russell, J. B. (1992). Another explanation for the toxic-iry of fermentation acids at low pH: anion accumulationversus uncoupling. J Appl Bacteriol 73, 363-370.

80. Saono, S., Hull, R. R. & Dhamcharee, B., eds. (1986). AConcise Handbook of Indigenous Fermented Foods inthe ASCA Countries. Canberra, Australia: The Govern-ment of Australia.

81. Simango, C. & Rukure, G. (1991). Survival ofCampylobacter jejuni and Escherichia coli in mahewu,

a fermented cereal gruel. Trans Royal Soc MedHyg 85,399^00.

82. Smith, J. E. & Solomons, G. L. (1994). Mycotoxins inhuman nutrition and health. DG XIL Brussels, Belgium:European Commission.

83. Sorrells, K. M., Enigl, D. C. & Hatfield, J. R. (1989).The effect of pH, acidulant, time and temperature on thegrowth and survival ofListeria monocytogenes. J FoodProt 52, 571-573.

84. Speck, M. L., Dobrogosz, W. J. & Casas, I. A. (1993).Lactobacillus reuteri in food supplementation. FoodTechnol 47(7), 90-94.

85. Stevens, K. A., Sheldon, B. W., Klapes, N. A. &Klaenhammer T. R. (1991). Nisin treatment for inacti-vation of Salmonella species and other gram-negativebacteria. Appl Environ Microbiol 57, 3613-3615.

86. Stevens, K. A., Sheldon, B. W., Klapes, N. A. &Klaenhammer, T. R. (1992). Effect of treatment condi-tions on nisin inactivation of gram negative bacteria. /Food Prot 55, 763-767.

87. Svanberg, U., Sjgren, E., Lorri, W., Svennerholm, A. -M. & Kjaiser, B. (1992). Inhibited growth of commonenteropathogenic bacteria in lactic-fermented cerealgruel. World J Microbiol BiotechnolS, 601-606.

88. Tsai, Y. W. & Ingham, S. C. (1997). Survival of E. coliO157:H7 and Salmonella spp. in acidic condiments. JFood Prot 60, 751-755.

89. Vasconcelos, A. T., Twiddy, D. R., Westby, A. &Reilly, P. J. A. (1990). Detoxification of cassava duringgari preparation. Int Journal of Food Sd Technol 25,198-203.

90. Verheul, A., Russell, N. J., Vant Hof, R., Rombouts, F.M. & Abee, T. (1997). Modifications of membranephospholipid composition in nisin-resistant Listeriamonocytogenes Scott A. Appl Environ Microbiol 63,3451-3457.

91. Westby A. & Choo, B. K. (1994). Cyanogen reductionduring lactic fermentation of cassava. Acta Hort 375,209-215.

92. Wheater, D. M., Hisch, A. & Mattic, A. T. R. (1952).Possible identity of "lactobacillin" with hydrogen per-oxide produced by lactobacilli. Nature 70, 623-626.

93. Yang, Z., Suomalainen, T., Mayra-Makinen, A. &Huttunen, E. (1997). Antimicrobial activity of 2-pyrrolidone-5-carboxylic acid produced by lactic acidbacteria. J Food Prot 60, 786-790.

94. Yokotsuka, T. & Sasaki, M. (1998). Fermented proteinfoods in the Orient: shoyu and miso in Japan. In Micro-biology of Fermented Foods, 2nd edn, pp. 351-415. Ed-ited by B. J. B. Wood. London: Blackie Academic andProfessional.

95. Yusof, R. M., Morgan, J. B. & Adams, M. R. (1993).Bacteriological safety of a fermented weaning food con-taining L-lactate and nisin. J Food Prot 56, 414-417.

CHAPTER 3

An Introduction to the HazardAnalysis and Critical Control Point

(HACCP) System and Its Applicationto Fermented Foods

Yasmine Motarjemi

INTRODUCTION

The Hazard Analysis and Critical ControlPoint system, often known under its acronymHACCP, is a system that was conceived in the1960s by the Pillsbury Company, NationalAeronautics and Space Administration (NASA),and the U.S. Army Laboratories at Natick to en-sure the safety of foods for astronauts. Origi-nally developed to ensure the microbiologicalsafety of foodstuffs, it is now recognized that thesystem can be equally applied to control and pre-vent chemical and physical hazards.11

In the 30 years since its conception, theHACCP system has grown to become the uni-versally recognized and accepted method forfood safety assurance. In appreciation of the im-portance of the HACCP system for enhancingfood safety, and in view of the importance ofglobalization of the food supply and the need forharmonization of food safety requirements, theCodex Alimentarius Commission (CAC)adopted in 1993 the Guidelines for the Applica-tion of HACCP, which have received interna-tional recognition. In 1997, on the recommenda-tion of the World Health Organization (WHO),the CAC guidelines were revised and improved.5

For more than 20 years, WHO has recognizedthe importance of the HACCP system for the pre-vention of food-borne diseases and has played animportant role in its development, promotion, andharmonization. One of the most important initia-tives of the organization has been the promotion ofthe HACCP system in small industries and cottage

industries and the use of the HACCP system forhealth education activities.

Before explaining the HACCP system and itsapplication to the preparation of fermentedfoods, it is essential to understand why the sys-tem has received such recognition by the foodindustry and has been promoted by public healthauthorities, in particular WHO. For this purpose,the historical development of food safety needsto be reviewed.*

HISTORICAL DEVELOPMENT

Food safety has been of concern to human-kind since the dawn of history, and many of theproblems encountered in our food supply goback to the earliest recorded years. Many rulesand recommendations advocated in religious orhistorical texts are evidence of the concern toprotect people against food-borne hazards andfood adulteration. Although advances in scienceand technology in the last few centuries have in-creased scientists' understanding of chemicaland biological hazards, up until the early 1980s,the food hygiene1^ systems in many countrieswere based on empirical knowledge acquiredthrough the surveillance of food-borne diseases

*Food safety is the assurance that food will notcause harm to the consumer when it is prepared and/oreaten according to its intended use.

fFood hygiene is defined as all conditions and mea-sures necessary to ensure the safety and suitability offood at all stages of the food chain.

and, in some instances, on the perception ofwhat is considered a hazard. Often, this percep-tion led to the fact that food hygiene was inter-preted simply as cleanliness.

Because of the prevalence of diseases trans-mitted by the fecal-oral route, such as typhoidfever and cholera, the emphasis of food safetyprograms for a long time was on the improve-ment of water supply and sanitation infrastruc-tures and the protection of food from fecal con-tamination. As a result, regulatory authoritiesfocused their inspection activities on the clean-ing and sanitation of food businesses and on thepersonal hygiene of food handlers. A similar ap-proach was adopted in health education activi-ties, and the focus of food hygiene educationprograms was based mainly on hand washing,boiling of water, and the protection of foodfrom flies. These measures, where imple-mented, have been effective in reducing the in-cidence of fecal-orally transmitted food-borneand waterborne diseases, such as typhoid feverand cholera. However, they have been insuffi-cient to prevent other types of food-borne dis-eases, such as salmonellosis and campylobac-teriosis. With the recent drastic changes in foodproduction and lifestyles, some of these dis-eases have even increased in incidence. In addi-tion, other types of problems have emerged.6

THE NEED FOR CHANGE

The end of the twentieth century was markedby significant changes in the food safety assur-ance system and food control. Traditionally, themethod of food safety assurance was based ontwo types of measure: (1) actions undertakenduring the procurement of raw materials, pro-cessing, manufacturing, transport, and distribu-tion, including design, layout, and cleaning ofpremises to prevent contamination; and (2) ac-tions undertaken to ensure that food, which wasproduced, was indeed safe. The former wereusually prescribed in the Codes of Good Manu-facturing Practice (GMP) and/or the Codes ofHygienic Practice (GHP). For the purpose of thelatter, industries tested the end product for con-firmation of safety. Food control and public

health authorities also inspected premises forcompliance with GMP/GHP and other regula-tory requirements. As mentioned above, theseinspections often focused on the cleanliness ofpremises, food handlers, and the immediate en-vironment and failed to identify shortcomings re-lated to the production and preparation procedureitself. Authorities also carried out independenttesting of end products. These methods of foodsafety assurance and food control showed certainweaknesses: The control of premises was based onrandom inspection and not on what happened dur-ing longer periods of time before or after the in-spection. End-product testing performed by indus-try or food inspectors also proved to be costly andtime consuming and provided insufficient assur-ance of food safety (see the following section onthe benefits of the HACCP system).

Concomitant with the recognition of the limi-tations of traditional approaches to food safetycontrol, the concern for food safety grew. Therewere many reasons for these concerns.3

• increasing incidence of food-borne dis-eases in many parts of the world

• recognition that one of the major healthproblems of developing countries, infantdiarrhea, is to a large extent food-borne

• emergence of new or newly recognizedfood-borne pathogens such as verocyto-toxin-producing Escherichia coll, Campy-lobacter spp., food-borne trematodes, andSalmonella enteritidis

• increased knowledge and awareness of theserious and chronic health effects of food-borne hazards

• increased number of vulnerable people,such as the elderly, children, pregnantwomen, immunocompromised individuals,the undernourished, and individuals withother underlying health problems

• increased awareness of the economic con-sequences of food-borne diseases

• the possibility of detecting minute amountsof contaminants in food, due to advances inscientific and analytical methods

• industrialization and mass production lead-ing to (1) increased risks of large-scale food

contamination and (2) the considerablylarger numbers of people affected in food-borne disease outbreaks as a result

• urbanization, leading to a longer and morecomplex food chain, and thus greater possi-bilities for food contamination

• new food technologies and processingmethods, causing concern either about thesafety of the products themselves or aboutthe eventual consequences due to inappro-priate handling in households or food ser-vice and catering establishments

• changing lifestyles, depicted by an increas-ing number of meals eaten out of the home,either in food service and catering estab-lishments or at street food stalls

• increased tourism worldwide and interna-tional trade in foodstuffs, both leading to agreater exposure to foreign and unfamiliarfood-borne hazards

• increased contamination of the environment• increased consumer awareness of food

safety• lack of or decreasing resources for food

safety

As a result of the above, but also recognizingthe limitations of the traditional approaches, pub-lic health authorities and the food industry haveboth recognized the need to move to a more pre-ventive, scientific, and cost-effective food safetyassurance approach, namely, the HACCP system.

WHAT IS THE HACCP SYSTEM?

The HACCP system is defined by the CAC asa system which identifies, evaluates, and con-trols hazards which are significant for foodsafety. The value of the HACCP system lies inthe fact that it is a scientific, rational, and sys-tematic approach to the identification, assess-ment, and control of hazards during production,processing, manufacturing, preparation, and useof food to ensure that food is safe when con-sumed. The HACCP system is based on sevenprinciples, as introduced in Appendix 3-A.2

Through the application of these principles,industries or food establishments that apply the

system will be led to review their process of foodproduction or preparation critically, to identifyhazards and control measures, to establish objec-tive criteria for ensuring safety, and to monitorthat control measures have been properly imple-mented, in particular at those steps during thefood production that are critical for food safety.They will also foresee the necessary correctiveactions when monitoring procedures indicatefailures in the control of hazards. Through veri-fication and documentation, they can also, at alltimes, verify the adequacy of their measures.Guidance for the application of these principleshas also been provided by the CAC text onHACCP Principles and Guidelines for its Appli-cation (Appendix 3-A). It is, however, recog-nized that for the application of the HACCP sys-tem to be successful, some additional measuresneed to be considered.

1. Management commitment. The successfulimplementation of the HACCP system re-quires a change in attitude of policy anddecision makers in all sectors concernedas well as management commitment. Al-though implementation of the HACCPsystem brings many benefits, and in thelong term may reduce financial costs, itsimplementation in the initial stages re-quires additional resources encompassingqualified personnel, technical support fa-cilities, equipment, training, and so forth.Such conditions can be met only wherethere is a management commitment.

2. Prerequisites for HACCP. Prior to apply-ing the HACCP system to any sector of thefood chain, that sector should be operatingaccording to the CAC General Principlesof Food Hygiene, the appropriate CACCodes of Practice, and appropriate foodsafety legislation.

3. Training. Adequate training of personnelis a key to effective implementation of theHACCP system. Such training should in-clude an explanation of the HACCP sys-tem, the reasons for and the objectives ofits application, as well as the responsibili-ties of each person involved in the imple-

mentation of the HACCP plan. The tasksof operators working at each critical con-trol point (CCP) should also be clearly de-fined and the operators should be trainedin performing them.2'11

BENEFITS OF THE HACCP SYSTEM

The benefits of the HACCP system are sum-marized in the following paragraphs.11

• The HACCP system overcomes many ofthe limitations of the traditional approachesto food safety control (generally based on"snap-shot" inspection and end-producttesting), includinga. limitations of "snap-shot" inspection

techniques in predicting potential foodsafety problems

b. the difficulty of collecting and examin-ing sufficient samples to obtain mean-ingful, representative information in atimely manner and without the high costof end-product analysis

c. reduction of the potential for product re-call

d. identification of problems without un-derstanding the causes

• The HACCP system allows for the identifi-cation of all conceivable, reasonably ex-pected hazards, even where failures havenot previously been experienced. It is there-fore particularly useful for new operations.

• The HACCP system is sufficiently flexibleto accommodate any changes that might beintroduced, such as progress in equipmentdesign, improvements in processing proce-dures, and technological developments re-lated to the product.

• The HACCP system will help target/directresources to the most critical part of thefood operation.

• With the HACCP system, one can expectan improvement in the relationship be-tween (1) food processors and food inspec-tors and (2) food processors and consum-ers. The HACCP system provides ascientifically sound basis for demonstrating

that all reasonable precautions have beentaken to prevent a hazard from reaching theconsumer. In this way, it encourages confi-dence in the safety of food products andthus promotes both confidence in the foodindustry and stability of food businesses.

• Data collected facilitate the work of foodinspectors for auditing purposes.

• The HACCP system is applicable to thewhole food chain, from the raw material tothe end product (i.e., growing, harvesting,processing or manufacturing, transport anddistribution, preparation, and consumption).

• The application of the HACCP system canpromote international trade by increasingconfidence in food safety.

• The HACCP system can be readily inte-grated into quality management systemssuch as total quality management, ISO9000, and so forth.

AREAS OF APPLICATION

The HACCP principles can be applied in avariety of way s.

• The HACCP system is a system that is usedas a method of food safety assurance infood production, processing, manufactur-ing, and preparation. The CAC guidelinesfor the application of the HACCP systemprovide guidance on how the seven prin-ciples of HACCP can be implemented infood industries in order to have the greatestchance of success (Appendix 3-A).

• The HACCP system is amenable to effec-tive food control. It allows for more effi-cient inspection of food operations becausethe role of food inspectors is centered onthe assessment of the HACCP plan andconfirmation that it is designed properlyand operating effectively.

• The HACCP principles, in particular prin-ciples 1 to 5, can be used to study food prepa-ration practices and to identify and assesshazardous behavior, which should be thefocus of health education interventions.

• The HACCP concept can be used in the man-agement of overall food safety programs toidentify those problems all along the foodchain that are of greatest risk to public health,and to prioritize interventions.

THE HACCP SYSTEM IN FOODHYGIENE

Today, to achieve food safety, it is recognizedthat there is a need to apply measures of in-creased specificity (Figure 3-1). At a more gen-eral level, the CAC outlines the general prin-ciples of food hygiene. These principles lay thefoundation for food hygiene. Second, moreproduct-specific hygienic measures may be ap-plied to focus better on issues that are relevant tospecific commodities. These measures, also pre-scribed by the CAC, are described in specificcodes of manufacturing or hygienic practices.The CAC has developed codes for a number ofproducts, such as smoked or salted fish, curedham, and so forth. Finally, application of theHACCP system can further enhance food safetyby providing a mechanism for analyzing the haz-ards for each food or process, developing a tai-lor-made plan for ensuring food safety with em-phasis on CCPs, and ensuring that the criticallimits at these points are met. With each layer of

the above measures, the degree of assurance offood safety increases.4

To harmonize and promote the application ofthe HACCP system in food industries, the CACoutlined the principles and elaborated the guide-lines for the application of the system. The prin-ciples of the HACCP system, as defined by theCAC, set the basis for the minimum require-ments for mandatory application of the HACCPsystem. The guidelines are, on the other hand, ageneral guidance, and adherence to them is vol-untary. The CAC text on the HACCP Principlesand Guidelines for its Application are presentlyannexed to the CAC General Principles of FoodHygiene, and their application is consequentlyrecommended. However, due to the status of theCAC in international trade in food, the applica-tion of the HACCP system to the production,processing, or manufacturing of food may insome countries become compulsory for food ex-port. The reason is that since the conclusion ofthe GATT Uruguay Round of Multilateral TradeNegotiations in April 1994 and the coming intoforce of the World Trade Organization (WTO)Agreement on Sanitary and Phytosanitary Mea-sures, the work of CAC is recognized as the ref-erence or "yardstick" for national food safety re-quirements. As a result, members of the WTOneed to take the work of CAC into consideration

Figure 3-1 HACCP in food hygiene: Each additional measure of food hygiene will increase the degree of foodsafety assurance.14

and adapt their legislation to the provisions pro-vided by CAC. In recent years, some countriessuch as the United States and the EuropeanUnion have made application of the HACCPsystem compulsory in the production and pro-cessing of certain foodstuffs such as seafood, in-cluding those that are imported.

The implementation of the HACCP system inall food industries is an established goal in manycountries. However, most progress made inimplementing the HACCP system so far hasbeen achieved in medium- and large-scale foodindustries, particularly in industrialized coun-tries. Worldwide, analyses of food-borne dis-ease outbreaks show that the greatest majority offood-borne disease outbreaks result from mal-practices during food preparation in small busi-nesses, canteens, and homes. In small businessesand homes in both developed and developingcountries, the application of the HACCP systemmeets with greater difficulties.

In recognition of the need for improving thesafety of foods that are prepared or processed inhomes, food service establishments, street foodvendors, and "cottage" industries, WHO pro-posed the use of HACCP systems to small op-erations and for health education purposes.1'7'10

APPLICATION OF THE HACCPAPPROACH TO FOOD PREPARATION

The CAC General Principles of Food Hygieneprovide guidance on the basic general require-ments that are essential to ensuring food safety aswell as food suitability. WHO has adapted theseprinciples to different settings (e.g., food serviceestablishments, street food vendors, and house-holds) and recommended specific hygienic mea-sures for the preparation of food under such condi-tions.9'12'13 Although these food hygiene principlesare fundamental to achieving food safety, their ap-plication for ensuring safety nevertheless has cer-tain limitations.10

• They do not provide a mechanism for pri-oritizing control measures, even thoughsome measures may prove to be more criti-cal than others.

• The guidance is nonspecific to foods, op-erations, special socioeconomic conditionswhere the food is prepared, or the culturalfactors leading to high-risk practices.

Therefore, it is recommended that HACCPstudies be conducted in these settings to identifythe measures that are critical for ensuring thesafety of foods in the given sociocultural andeconomic conditions of preparation. The identi-fied measures should be considered CCPs and,depending on the circumstances, they should bethe subject either of enforcement and/or of train-ing and education.

In general, small food businesses or streetfood vendors do not have the expertise necessaryfor conducting such HACCP studies. The sameis also true for domestic food handlers. It is therole of trade associations, food inspectors or, inthe case of domestic food handlers, health au-thorities to conduct such studies and train oreducate the food businesses and households inimplementing and monitoring control measuresat the CCP in an adequate manner. Such an ap-proach provides a mechanism for prioritizingkey behavior or practices that should be en-forced or promoted in a given professional groupor population.

In most cases, a complete HACCP study can-not be conducted for every type of food andpreparation method, and priorities must, there-fore, be set. Whenever possible, epidemiologicaldata should be used in establishing priorities.High priority should be given to foods that arecommonly implicated as vehicles of food-bornediseases and to the type of food operationswhere outbreaks of food-borne diseases havebeen reported. Data on food-borne diseases are,however, not always available. In the absence ofsuch data, priorities can be set based on the fac-tors described in the following paragraphs.

Intrinsic Properties of the Food Involved

Some foods may contain toxic chemicals ormicrobial pathogens or their toxins because ofthe practices involved in the production of theraw materials. For example, raw meats may be

contaminated with microbial pathogens at theslaughtering stage, and raw vegetables may becontaminated with microbial pathogens or toxicchemicals from fertilizers, pesticides, and soforth. Food properties also relate primarily tocharacteristics of the food that may support thesurvival and/or growth of microorganisms basedon knowledge of microbial ecology and epide-miological history. The characteristics that aremost useful are pH, water activity (Aw), and re-dox potential (Eh). These factors influence thegrowth of infectious or toxigenic microorgan-isms. Foods that are possibly hazardous becausethey readily support rapid and progressivegrowth of microorganisms should be given highpriority. Next, priority should be given to foodsthat can support the growth of pathogens duringprolonged storage periods. Foods that are "rela-tively stable," in particular with regard to patho-genic organisms, can be assigned lower priority.

Preparation and Handling

Food operations that commonly contribute tothe cause of food-borne illnesses are those that (1)prepare hazardous foods in advance of serving, (2)store foods in a manner that might allow microbialgrowth, and (3) inadequately reheat food to inacti-vate pathogens or toxins. On the other hand, foodthat is thoroughly cooked just before consumptionis safe from biological hazards although chemicalsand certain toxins may not be affected. Food thathas been processed, even in a simple form such asfermentation, may be safe when held at ambienttemperatures. Similarly, commercially processedfoods, especially those that are well packaged,may pose little hazard to the consumer when theyare sold by street vendors.

Volume of Food Prepared

The concern about volume of food preparedrelates primarily to the amount of food preparedin advance of sale and/or consumption. For in-stance, in street- vending operations, this amountcan be indirectly measured by the average dailysales, the amount of prepared foods on display,

and the duration of holding cooked foods ondisplay.

Susceptibility of Consumers

Infants and children, pregnant women, thehospitalized, immunocompromised persons, andthe elderly are more susceptible to food-bornediseases than the general population. Foodswhich are intended for these groups of consum-ers should receive a higher priority.

Many of the fermented foods, particularly thosethat are prepared as complementary foods for in-fants and young children in Africa, fall into one ormore of the above categories. In view of this,WHO joined with the Food and Agricultural Or-ganization (FAO) in December 1995 to conduct aworkshop on the assessment of fermentation as ahousehold technology to improve food safety,during which the HACCP system was applied toseveral African fermented foods.8 CCPs for thepreparation of each of these foods were deter-mined and it was recommended that control mea-sures at these points should be included in thehealth education of food handlers. As an example,the application of the HACCP system to the prepa-ration of gari is presented below. HACCP studieshave shown that in general, the following mea-sures during the preparation of African fermentedfoods should be the focus of training and healtheducation.

• rapid and adequate acidification• sufficient cooking for rendering food safe• sufficient size reduction of cassava to en-

able enzymatic degradation of toxic cyano-genic glycosides

• avoidance of moldy raw material in view ofpossible mycotoxin contamination

• use of safe water

Through these studies, it could also be clearlydemonstrated that some preparation practicespresented a high food safety risk. For instance,the preparation of togwa (see Chapter 12) in-volves the addition of "power flour" after thefood has been cooked. Thus, there is a risk ofpostcooking contamination, particularly with

acid-resistant pathogens. In addition, if the lacticacid fermentation fails, the addition of powerflour may lead to the proliferation of bacterialpathogens. To minimize food safety risks, it wasrecommended to accelerate the fermentation byback-slopping or use of starter cultures, or usingpower flour of high quality and avoidingpostcooking contamination. In addition, theHACCP study showed that the original proce-dure for power flour preparation needs to bemodified. In order to avoid postcooking con-tamination, it was recommended to boil the wa-ter that was used for the purpose of soaking andgermination.

APPLICATION OF THE HACCP SYSTEMTO GARI*

1. Product description — Gari is a granularstarchy food that is made from cassavaroots. The processing starts with peeling,washing, and grating the tubers. Thegrated pulp is then put into bags (often juteor woven polypropylene bags) and left toferment for several days under weight(pressure), during which time water is alsoremoved. Fermentation is followed byfragmentation, drying, and roasting. Dur-ing the roasting stage, the core tempera-ture reaches 80-85 0C and the starch is ge-latinized. Palm oil is sometimes addedduring roasting. After the roasting pro-cess, the gari, as it is now called, is cooledand stored. The final moisture content willdetermine its shelf life. When a final mois-ture content of less than 10% is reached,gari may be stored for several months. Athigher moisture contents, the shelf life ofgari is reduced to a few weeks because ofpotential mold growth.

* Application of the HACCP system has been sim-plified and adapted to household conditions. Al-though the same approach can be used for productionon a cottage and industrial scale, the requirements interms of CCPs, critical limits, and monitoring proce-dures may be different and more severe. ModelHACCP plans are not appropriate for use until theyare validated for a specific food and food process.

2. Intended use — Gari is an important partof the staple diet in Nigeria and manyother African countries. It is also given tochildren over one year of age. Gari can beprepared in many different ways. In thefollowing example, soaking in cold wateris used.

3 . Flow diagram — The flow diagram of gariis presented in Figure 3-2.

4. Hazards of concern — Hazards consid-ered in this context include biological(e.g., bacteria, viruses, parasites), chemi-cal (e.g., contaminants, mycotoxins), andphysical agents.

5. Identification of hazards, control mea-sures, and CCPs- Table 3-1 shows thehazards that are associated with each stepin the preparation of gari and some pos-sible measures for their control,a. Raw material: Major hazards of cassava

are cyanogenic gluco sides (i.e.,linamarin and lotaustralin) and contami-nation by agrochemicals. Cyanogenicglucosides will be hydrolyzed and re-moved during later stages in the process-ing and preparation of gari (i.e., grating,fermenting, and roasting). However, inregard to chemical contaminants andagrochemicals, households should ob-tain assurance from the suppliers regard-ing the safety of raw products.

Depending on its source, water maybe contaminated. Although the micro-biological safety of water is of lesserimportance when washing cassava be-cause it will be fermented and heattreated, it is critical that safe water isused in the final preparatory stages be-fore consumption,

b. Peeling: Some foreign matter andpathogens may be introduced at thisstep. However, because the cassavawill be washed and heat treated, thecontrol of hazards, other than peeling inhygienic conditions, is not critical atthis stage,

c. Washing: Microbiological hazardsmay be introduced if the water is notclean. Therefore, as part of a good hy-

Washing

Grating

Bagging

Fermentation(in bags under weight)

Roasting

Cooling

Storage

Cassava

Peeling

Water

Soaking

Water

Serving

Figure 3-2 Application of the HACCP system to the preparation ofgari in households8

gienic practice, safe water should beused. Hazards introduced at this stepcan nevertheless be controlled duringsubsequent steps of gari production.Washing can, though, decrease theamount of foreign matter.

d. Grating: This is the most importantstep with regard to detoxification whenthe cellular disruption results in the re-lease of linamarase enzymes andgreater contact of the enzymes with itssubstrate linamarin. Therefore, grating

Table 3-1 Application of the HACCP System to the Preparation of Gar/ in Households

Corrective Actions

a. Use anothersource of water

b. i) Boil the water

ii) Reboil

b. Reclean

continues

MonitoringProcedure

a. Observation,smelling, andtasting

b. i) Inquiry fromhealth authorities

ii) Observation

b. Observation

Critical Limit

a. Clear, free of odorand off-taste

b. i) Indication ofcontamination

ii) Bubbles

b. As clean aspossible

Critical ControlPoints

a. No

b. No

c. No

a. Yes

b. Yes for step 1 0

a. Nob. No

a. No

b. Yes

Control Measures

a. Obtain assurancefrom supplier ofadequatepreharvest andpostharvesthandling of roots

b. Grating, fermenta-tion, and roasting

c. Heat treatment

a. Obtain assuranceof the source ofwater; use onlysafe water

b. i) Use safe water(i.e., filtered anddisinfected) orii) Boil the water

a. Washingb. Heat treatment

a. Use safe water

b. Thorough washing

Hazards

a. Agrochemicals

b. Cyanogenicglucosides

c. Pathogens

a. Chemicalcontaminants,depending on thesource

b. Pathogens (e.g.,pathogenic E.coli,Campylobacter, V.cholerae,Salmonella,Cryptosporidium,Giardia lamblia,Entamoebahistolytica),Rotavirus

a. Foreign matterb. Pathogens

a. Introduction ofpathogens throughwater

b. Residual foreignmatter

Step

1 . Raw materiali) Cassava

1 . Raw materialii) Water

2. Peeling

3. Washing

Table 3-1 continued

Corrective Actions

a. Reg rate

c. Remove foreignmaterial

Use other bags

Discard

a. i) Continueroastingii) Breaking uplumps

b. Same as above

continues

MonitoringProcedure

a. Observation

c. Observation

Monitor the sourceand other uses ofbags

Observation,smelling, and tasting

a. i)Time keeping

ii) Observation

b. Same as above

Critical Limit

a. Absence of coarseparticles

c. Free of visibleforeign matter

Absence of chemicalcontaminants

Acid taste andcharacteristic odorswithin 24 hours

a. i) Sufficient timefor roastingii) Small sizeparticles

b. Same as above


a. Yes

b. No

c. Yes

Yes

Yes

a. Yes

b. Yes

Control Measures

a. Complete grating

b. Clean theequipment

c. Use well-maintainedequipment

Use clean bags;obtain assurancefrom supplier aboutno previous andhazardous use of thebags

Rapid fermentation

a. i) Thoroughroastingii) Breaking uplumps

b. Same as above

Hazards

a. Residual cyanide

b. Pathogens

c. Foreign matter

Chemicalcontamination

Growth of pathogensand production oftoxin (e.g., Staphylo-coccus aureus)

a. Residual cyanide

b. Mold growthduring storage ifhigh moisturecontent

Step

4. Grating

5. Bagging

6. Fermentationunder weight

7. Roasting

Table 3-1 continued

Corrective Actions

a. See step 1 (water)

b. Thorough heating

c. Thorough heating

MonitoringProcedure


b. Observation

c. Time keeping

Critical Limit


b. Washing with soapand thoroughrinsing with cleanwater

c. Use within fourhours


No

No, see step 7

a. Yes see step 1

b. Yes

c. Yes

Control Measures

Cool under hygienicconditions (e.g., put inclean container andclean environment)

Thorough roasting(see step 7:roasting); keep in dryconditions

a. Use safe water

b. Wash hands anduse clean utensils

c. Consumptionwithout delay

Hazards

Contaminationthrough environment

Mold growth duringstorage if highmoisture content

a. Recontaminationwith water

b. Recontaminationby dirty hands,utensils, environ-ment

c. Growth ofpathogens andspores of Bacilluscereus, ifconsumption isdelayed for morethan four hours

Step

8. Cooling

9. Storing

10. Soaking andserving

must be thorough to ensure a fast deg-radation of linamarin.

Foreign matter and pathogens mayalso be introduced at this stage. Al-though pathogens can be killed at theroasting step, the prevention and elimi-nation of any foreign matter at this stepis essential,

e. Bagging: Unclean bags may furthercontaminate the raw material. Chemi-cal contamination is of particular con-cern at this step. The bags should nothave been used previously for purposesthat could jeopardize the safety of gari(e.g., for storage of pesticides),

f. Fermentation: Rapid fermentation isimportant to prevent the growth of un-desirable microorganisms and the pro-duction of toxins. Fermentation istherefore a CCP for the control ofpathogens.

Fermentation also provides the op-portunity (contact time) necessary forthe action of linamarase on its sub-strate. During later stages, however,fermentation may have an antagonisticeffect on detoxification because the de-crease in pH resulting from fermenta-

REFERENCES

1. Bryan, F. L. (1992). Hazard Analysis Critical ControlPoint Evaluations: A Guide to Identifying Hazards andAssessing Risks Associated with Food Preparation andStorage. Geneva, Switzerland: World Health Organiza-tion.

2. Codex Alimentarius Commission. (1997). Food Hy-giene, Basic Texts. Secretariat of the Joint FAOAVHOFood Standards Program. Rome: Food and AgriculturalOrganization.

3. Motarjemi, Y. & Kaferstein, F. (1999). Food safety,HACCP and the increase in foodborne diseases: a para-dox? Food Control 10, 325-333.

4. Motarjemi, Y. & Van Schothorst, M. (1999). HACCP:Principles and Practice. Geneva, Switzerland: WorldHealth Organization.

5. Motarjemi, Y., Kaferstein, F., Moy, G., Miyagawa, S. &Miyagishima, K. (1996). Importance of HACCP for

tion may lead to the stability of cyano-hydrins. An optimization of the fer-mentation process with respect to hy-drolysis of linamarin and control ofmicrobial growth is therefore importantfor ensuring the chemical and micro-biological safety of gari.

g. Roasting: Further detoxification of cas-sava occurs during roasting when thehydrogen cyanide is evaporated. Thor-ough drying at this step is also impor-tant for the stability of gari and the pre-vention of mold growth during storage.It is important to prevent lumps fromforming because they may limit thedrying and evaporation of hydrogencyanide,

h. Cooling: Cooling should take place un-der hygienic conditions,

i. Storing: To prevent mold growth, garishould be kept under dry conditionsand protected from animals and ro-dents,

j. Serving: Water used for soaking gari, aswell as hands and utensils, may reintro-duce pathogens. It is critical that the wa-ter used at this step is safe, and that uten-sils and hands are washed thoroughly.

public health and development: the role of the WorldHealth Organization. Food Control 7, 77-85.

6. Motarjemi, Y. (1999). The starting point: what is foodhygiene? New Food 2, 25-30.

7. World Health Organization. (1993). Application of theHazard Analysis Critical Control Point (HACCP) Sys-tem for the Improvement of Food Safety: WHO-Sup-ported Case Studies on Food Prepared in Homes, atStreet Vending Operations, and in Cottage Industries.Unpublished document WHO/FNU/FOS/93.1. Geneva,Switzerland: Author.

8. World Health Organization. (1996). Assessment of Fer-mentation as a Household Technology for ImprovingFood Safety: Report of a WHO Consultation. Unpub-lished document WHO/FNU/FOS/96.1. Geneva, Swit-zerland: Author.

9. World Health Organization. (1996). Basic Principles forthe Preparation of Safe Food for Infants and YoungChildren. Unpublished document WHO/FNU/FOS/96.6. Geneva, Switzerland: Author.

10. World Health Organization. (1996). Essential SafetyRequirements for Street-Vended Foods, rev. edn. Un-published document WHO/FNU/FOS/96.7. Geneva,Switzerland: World Health Organization.

11. World Health Organization. (1997). HACCP: Introduc-ing the Hazard Analysis Critical Control Point System.

WHO document WHO/FNU/FOS/97.2. Geneva, Swit-zerland: World Health Organization.

12. World Health Organization. The WHO Golden Rules forSafe Food Preparation. Unpublished document.Geneva, Switzerland: World Health Organization.

13. World Health Organization/BGW. (1994). Hygiene inFood-Service and Mass Catering Establishments. Un-published document WHO/FNU/FOS/94.5. Geneva,Switzerland: World Health Organization.

APPENDIX 3-A

HACCP Principles andGuidelines for Its Application

Assemble HACCP team

The food operation should ensure that the appropriate product-specific knowledge and expertise isavailable for the development of an effective HACCP plan. Optimally, this may be accomplished byassembling a multidisciplinary team. Where such expertise is not available on-site, expert adviceshould be obtained from other sources. The scope of the HACCP plan should be identified. The scopeshould describe which segment of the food chain is involved and the general classes of hazards to beaddressed (e.g., does it cover all classes of hazards or only selected classes).

Describe product

A full description of the product should be drawn up, including relevant safety information such ascomposition, physical/chemical structure (including Aw, pH, etc.), microbicidal/static treatments (e.g.,heat treatment, freezing, brining, smoking, etc.), packaging, durability, storage conditions, and methodof distribution.

Identify intended use

The intended use should be based on the expected uses of the product by the end user or consumer.In specific cases, vulnerable groups of the population such as institutional feeding may have to beconsidered.

Construct flow diagram

The flow diagram should be constructed by the HACCP team. The flow diagram should cover allsteps in the operation. When applying HACCP to a given operation, consideration should be given tosteps preceding and following the specified operation.

On-site confirmation of flow diagram

The HACCP team should confirm the processing operation against the flow diagram during allstages and hours of operation and amend the flow diagram where appropriate.

HACCP principles

Principle 1Conduct a hazard analysis

Principle 2Determine the criticalcontrol points (CCPs).

Principle 3Establish critical limit(s).

Guidelines for their application

The HACCP team should list all of the hazards that may bereasonably expected to occur at each step from primary production,processing, manufacture, and distribution until the point of con-sumption.

The HACCP team should next conduct a hazard analysis to identifyfor the HACCP plan which hazards are of such a nature that theirelimination or reduction to acceptable levels is essential to theproduction of a safe food.

In conducting the hazard analysis, wherever possible, the followingshould be included:

• the likely occurrence of hazards and severity of their adversehealth effects

• the qualitative and/or quantitative evaluation of the presence ofhazards

• survival or multiplication of microorganisms of concern• production or persistence in foods of toxins, chemicals, or

physical agents• conditions leading to the above

The team must then consider what control measures, if any, existthat can be applied for each hazard.

More than one control measure may be required to control aspecific hazard(s) and more than one hazard may be controlled bya specified control measure.

There may be more than one CCP at which control is applied toaddress the same hazard. The determination of a CCP in theHACCP system can be facilitated by the application of a decisiontree (Figure 3-A-1) which indicates a logic reasoning approach.Application of a decision tree should be flexible, given whether theoperation is for production, slaughter, processing, storage, distribu-tion, or other. It should be used for guidance when determiningCCPs. This example of a decision tree may not be applicable to allsituations. Other approaches may be used. Training in the applica-tion of the decision tree is recommended.

If a hazard has been identified at a step where control is necessaryfor safety, and no control measure exists at that step, or any other,then the product or process should be modified at that step, or atany earlier or later stage, to include a control measure.

Critical limits must be specified and validated if possible for eachcritical control point. In some cases, more than one critical limit willbe elaborated at a particular step. Criteria often used includemeasurements of temperature, time, moisture level, pH, Aw, andavailable chlorine, and sensory parameters such as visual appear-ance and texture.

Do control measure(s) exist?

Modify step,process, or product

Is control at this stepnecessary for safety?

Not a CCP Stop

Is the step specifically designed toeliminate or reduce the likely occurrence

of a hazard to an acceptable level?

Could contamination with identifiedhazard(s) occur in excess of acceptable

level(s) or could these increase tounacceptable levels?

Not a CCP Stop

Will a subsequent step eliminateidentified hazard(s) or reduce likelyoccurrence to an acceptable level?

CRITICAL CONTROL POINT

Not a CCP Stop

* Proceed to the next identified hazard in the described process.

Figure 3-A-l Example of a decision tree to identify CCPs.2

HACCP principles

Principle 4Establish a system tomonitor control of theCCPs.


Monitoring is the scheduled measurement or observation of a CCPrelative to its critical limits. The monitoring procedures must be ableto detect loss of control at the CCP. Further, monitoring shouldideally provide this information in time to make adjustments toensure control of the process to prevent violating the critical limits.Where possible, process adjustments should be made when

HACCP principles

Principle 5Establish the correctiveaction to be taken whenmonitoring indicates that aparticular CCP is not undercontrol.

Principle 6Establish procedures forverification to confirm thatthe HACCP system isworking effectively.

Principle 7Establish documentationconcerning all proceduresand records appropriate tothese principles and theirapplication.


monitoring results indicate a trend toward loss of control at a CCP.The adjustments should be taken before a deviation occurs. Dataderived from monitoring must be evaluated by a designated personwith knowledge and authority to carry out corrective actions whenindicated. If monitoring is not continuous, then the amount orfrequency of monitoring must be sufficient to guarantee that theCCP is in control. Most monitoring procedures for CCPs will need tobe done rapidly because they relate to on-line processes and therewill not be time for lengthy analytical testing. Physical and chemicalmeasurements are often preferred to microbiological testingbecause they may be done rapidly and can often indicate themicrobiological control of the product. All records and documentsassociated with monitoring CCPs must be signed by the person(s)doing the monitoring and by a responsible reviewing official(s) ofthe company.

Specific corrective actions must be developed for each CCP in theHACCP system in order to deal with deviations when they occur.

The actions must ensure that the CCP has been brought undercontrol. Actions taken must also include proper disposition of theaffected product. Deviation and product disposition proceduresmust be documented in the HACCP recordkeeping.

Establish procedures for verification. Verification and auditingmethods, procedures, and tests, including random sampling andanalysis, can be used to determine if the HACCP system is workingcorrectly. The frequency of verification should be sufficient toconfirm that the HACCP system is working effectively. Examples ofverification activities include

• review of the HACCP system and its records• review of deviations and product dispositions• confirmation that CCPs are kept under control

Where possible, validation activities should include actions toconfirm the efficacy of all elements of the HACCP plan.

Efficient and accurate recordkeeping is essential to the applicationof a HACCP system. HACCP procedures should be documented.Documentation and recordkeeping should be appropriate to thenature and size of the operation.

Documentation examples are:

• hazard analysis• CCP determination• critical limit determination

Record examples are:

• CCP monitoring activities• deviations and associated corrective actions• modifications to the HACCP system

CHAPTER 4

Chemical Hazards and Their Control:Endogenous Compounds

Leon Brimer

INTRODUCTION

Raw materials of vegetable origin may con-tain natural toxic or antinutritional compounds,endogenous constituents that are synthesized bythe plant itself. Antinutritional means a deleteri-ous effect due to the hindrance of uptake or useof other components in the diet. Examples ofantinutritional compounds include tannins,which among others bind to proteins, makingthem unacceptable as substrates for proteases;proteinase inhibitors, which inhibit proteinasessuch as trypsin and chymotrypsin; phytate, whichbinds a number of minerals, making them unavail-able for uptake; and thiaminase, which degradesvitamin B1. An effect may be due to the parentcompound or to metabolites that are formed in thegut or in the organism after absorption.

Although a few toxic and antinutritional com-pounds found in plants are proteins, most arelow molecular weight compounds. Examples ofproteins are ricin,53 which is found in the seedsofRicinus communis L. (castor bean); lectins,213

found especially within the legumes; and theproteinase inhibitors, which are also common inlegumes.62 The smaller molecules with deleteri-ous effects belong to the groups of compoundsthat are normally classified as secondary me-tabolites. The number of secondary constituentsisolated from plants, fungi, and animals is high.Luckner worked with 26 biosynthetic groups di-vided into 107 subgroups, many of which con-tained more than 1,000 structures.155 The major-ity of these compounds was found in plants. The

alkaloids, for example, have had the attention ofphytochemists for more than 150 years. In 1950,approximately 2,000 alkaloids were recognized;by 1970, the number had increased to approxi-mately 4,000; 20 years later, approximately10,000 were known.217 It is necessary, then, to fo-cus on the most important endogenous plant toxinsas seen from a food and feed point of view.

The most prominent constituents known to re-strict the nutritional value of food or fodder in-clude certain nonprotein amino acids, alkaloidsand glycosides, together with the tannins. How-ever, knowledge concerning the influence offer-mentation on these agents is very limited exceptfor certain of the glycosides. Because a numberof very important commodities of food and fod-der worldwide do contain toxic glycosides,46 theoccurrence and effects of toxic glycosides, andtheir fate during food fermentations, will be pre-sented in this chapter.

TOXIC AND ANTINUTRITIONALGLYCOSIDES IN FOOD AND FEED

Glycosides consist of one or more genins (ag-lycones) to which one or more mono- or oli-gosaccharides are linked. The glycosidiclinkage(s) may differ (i.e., one differentiates be-tween O-, S-, and C-glycosides) (Figure 4—1). Ifthe sugar part is a glucose moiety, it is called aglucoside. A number of different glycosides andoligosaccharides causing physiological effects(toxins) or reduced uptake or use of nutrients af-ter ingestion are known in the plant kingdom. So

Figure 4-1 (a) The general structure of O-, S-, and C-glucosides as representatives of the broader groups ofO-, S-, and C-glycosides, respectively. At top, an O-glucoside; middle, an S-glucoside; bottom, a C-glucoside.(b) Examples of naturally occurring O-, S-, and C-glucosides. Top (Linamarin—a cyanogenic glucoside found incassava), middle (Sinigrin—a glucosinolate), and bottom (Barbaloin—an anthrone C-glucoside from Aloe spp.;laxative). Note: R = aglycone (= genin).

are a number of bitter-tasting glycosides that re-duce the palatability of the plant (Table 4-1). Afew of these glycosides have been shown to beprotective to the plant;129'205 however, most haveonly been recognized as toxic to domestic animalsor humans. The broad range of compounds listedin Table 4-1 illustrates the diversity of chemicalstructures found even within the restricted field oftoxic and antinutritional glycosides and oligosac-charides. Because this diversity also means abroad range of different mechanisms of action, themost important compounds are described in thefollowing paragraphs.

Cyanogenic Glycosides

Cyanogenesis (cyano — Greek [kyanos =blue] and genesis — Greek [creation]) means theformation of cyanide/hydrogen cyanide, orHCN. Organisms that possess the ability to re-lease cyanide may be termed cyanogenic orcyanophoric (phoros — Greek [bearing]). If thecyanide is formed from the breakdown of an-other compound, this compound is called a cya-nogenic (cyanogenetic) compound, or simply, acyanogen. Cyanogens include cyanogenic gly-cosides, cyanogenic lipids, cyanohydrins, and

Table 4-1 Toxic, Antinutritional, and Bitter-Tasting Glycosides and Oligosaccharides

Compound group orcompound

Cyanogenic glycosides

Glycoalkaloids

Glycosides of organicnitriles

Glycosides and sugaresters of aliphaticnitrocompounds

Methylazoxymethanol(MAM) glycosides

Naringin

Oligosaccharides

Platyphylloside

Polyphenols

Ptaquiloside

Saponins

Vicine and convicine

Carboxyatractyloside(CAT) and relatedcompounds

Examples Toxicity, taste, etc.

O-Glycosides, Sugar Esters, and OligosaccharidesIn sources of food and feed

Linamarin in Manihot esculenta,Euphorbiaceae, in general wide-spread in the plant kingdom(Tracheophyta andSpermatophyta)52-129'182

Chachonin and solanin in Solanumtuberosum, Solanaceae(Angiospermae)140'206

Simmondsin in Simmondsia californica(Jojoba), Buxaceae(Angiospermae)2'277

Miserotoxin in Astragalus spp.,Fabaceae (= Leguminosae;Angiospermae)160'204

Cycasin in Cycasspp., Cycadaceae(Gymnospermae)22'154

In Citrus spp., espec. C. paradisi(grapefruit), Rutaceae(Angiospermae)227

In seeds of several legume spp.,Fabaceae ( = Leguminosae;Angiospermae)94'215

In Betula pendula, Betulaceae(Angiospermae)262

2-hydroxyarctiin in Carthamus tinctorius(Safflower), Asteraceae( = Compositae; Angiospermae)98'203

In Pteridium aquilinum, Polypodiaceae(Tracheophyta)249'250

Triterpene or steroid saponins inQuinoa spp., Borassus flabellifer,Glycyrrhlzae glabra and Balanitesspp. (Angiospermae)74'95'125

In Vicia faba (faba bean), Fabaceae( = Leguminosae; Angiospermae)276

In medicinal and toxic plants

CAT in Atractylis gummifera,Asteraceae (= Compositae;Angiospermae)49'187

Acute and chronic toxicity due torelease of HCN; neurotoxicityof intact glycosides dis-cussed; bitter taste

Corrosive to the gastrointestinaltract; upon absorption, acutelytoxic due to several mecha-nisms; bitter taste

Causes chronic toxicity ofunknown mechanism

Acutely toxic to ruminants;inhibit the TCA-cycle of thecells

Carcinogenic

Bitter taste

Flatulence-producing

Antinutritional (deterrrent) toseveral animal species

Cathartic (laxative); bitter taste

Acutely toxic and carcinogenic

Some atoxic, other mildly tostrongly toxic; several arebitter tasting

Acutely toxic to glucose-6-phosphate dehydrogenase-deficient individuals

Acutely toxic; inhibit mitochon-dria! oxidative phosphorylation

continues

Table 4-1 continued

Compound group orcompound

Cardeno- andbufodienolides

Cucurbitacins

Glycosides of Vitamin Ds

Ranunculin

Examples

"Digitalis" glycosides in Digitalis spp.(cardiac glycosides),Scrophulariaceae(Angiospermae)161'225

Cucurbitacin L in Citrullus colocynthis,Cucurbitaceae (Angiospermae),some cucurbitacins also present infood plants (ref. text below)103-126

Glycosides of 1oc,25-(OH)2D3 inSolanum glaucophyllum, Solanaceae(Angiospermae)279

In Ranunculus and Caratocephalusspp., Ranunculaceae(Angiospermae)180'195

Toxicity, taste, etc.

Acutely toxic to the heart

Intensely bitter substances,some of which are acutelytoxic

Chronic toxicity (vitamin Dintoxication — calcinoses)

Acutely toxic; irritant to mucousmembranes; Upon absorption,it affects several organs suchas the heart, the lungs, etc.

C-Glycosides (Some Also Occurring as O-Glycosides)In medicinal plants

Anthraquinone, an-throne, and dianthroneglycosides

Glucosinolates

Sennosides in Cassia angustifolia,Fabaceae (= Leguminosae;Angiospermae)152'280

S (Thio)-GlycosidesIn food and feed resources

In many species within the families ofCapparales (Angiospermae)23

Laxative effect; some com-pounds are drastica

Chronic toxicity due to releaseof thiocyanate and othercompounds; sharp (burning)taste

cyanogenic epoxides.31'182 Cyanogenesis has beendetected in prokaryotes, fungi, plants, and ani-mals. Cyanogens have been isolated from a greatnumber of organisms; the glycosides, however,have been isolated only from plants and insects.182

The release of cyanide from a cyanogen im-plies the degradation of the compound, a reac-tion that may be either spontaneous or enzymecatalyzed.52'182 Cyanogenic lipids and cyano-genic glycosides are broken down to cyanohy-drins (hydroxynitriles); these are cyanogens inthemselves (Figure 4-2). The cyanogenesisstarts on crushing of the tissue, the cyanogens,

and the degradative enzymes being compart-mentalized either at the subcellular level or attissue level in the intact plant.211

The glycosides are the most common cyano-gens, and comprise more than 60 structures.146'241

They were recognized early as substances that arepoisonous to animals.86 Cyanogens are of somesystematic importance at the level of higher planttaxa,182'265 and within certain families and gen-era 182,194,242 JJ16 ingestion of cyanide and cyano-genic compounds may lead to acute172 as well aschronic intoxications, the latter including the cen-tral nervous system (CNS) syndrome, konzo.269'270

2,3-Epoxynitrile Cyanogenic glycoside/lipid

Figure 4—2 The interrelationship between cyanogenic compounds and cyanide/hydrogen cyanide. In a cyano-genic glycoside, R1 is a saccharide moiety; in a cyanogenic lipid, an acyl moity. Hydrolases: glycosidase(s)—Refer also to Figure 4—3—or lipase. Note that the Cyanohydrin formed upon hydrolysis of one of the three typesof cyanogens (epoxynitriles, glycosides, or lipids) is a cyanogen itself.

Glycoalkaloids

Steroidal alkaloids and alkaloid glycosidesoccur throughout the genus Solanum (Solan-aceae). The common potato (S. tuberosum) con-tains in its edible tuber the two compounds a-chaconin and a-solanin.140 The total content mayvary from 10 mg/kg to 390 mg/kg, with a meanof 73 mg/kg.96 In other species of Solanum andclosely related genera, different glycosides andfree genins may dominate.92'206 Gastrointestinalabsorption of steroidal alkaloid glycosides var-ies between animal species. Some hydrolysis ofthe glycosidic bond and further metabolismseem to occur in different animal species, asjudged from analyses comparing the serum levelof oc-chaconin and/or a-solanin to that of totalalkaloids at different times after ingestion.140

The toxicity of the potato glycosides to humansincludes gastrointestinal upset with diarrhea,vomiting, and abdominal pain. In severe cases,neurological symptoms, some of which areclearly a result of the acetylcholinesterase in-

hibitor activity of these glycosides,216 are seen.140

Both a-chaconin and a-solanin, together withtheir aglycones, are teratogenic in one or moreanimal models.96 However, Kuiper-Goodman &Nawrot140 did not find the suggested associationof the consumption of blighted potatoes duringpregnancy with increasing incidences of spinabifida substantiated.

Methylazoxymethanol Glycosides

Glycosides of memylazoxymethanol (MAM)have been found only in cycads (Macrozamiaand Cycas spp). The concentration is high inseeds; smaller quantities are found in stems andleaves.154 Extensive losses of sheep have oc-curred in Australia due to consumption ofMacrozamia and Cycas spp.113 The first isola-tion of a MAM glycoside, macrozamin (the P-primeverosid of MAM), was from seeds of M.spiralis Miq., an Australian cycad. Today, otherMAM glycosides are known, among these cyca-sin (the (3-D-glucopyranosid of MAM), which

Cyanohydrin Cyanohydrin

Hydroxynitrilelyase or

Hydrogen Cyanide

Hydrolases Cyanide

was shown to be characteristic of, and exclusiveto, all the genera of cycads.154 The relative con-centrations of cycasin to macrozamin in ripeseeds differ within the cycad genera.177 The gly-cosides release MAM on the hydrolysis, whichis catalyzed by (3-glycosidases. MAM is a mu-tagenic and carcinogenic alkylating agent.163'178

Oligosaccharides

Flatulence is a common phenomenon that isassociated with the ingestion of legumes, amongothers, and caused by the microbial fermentationof low molecular weight sugars. Many of thesesugars are a-galactosides because humans donot have oc-galactosidase in their digestivetract.104 The legume oligosaccharides, raffmose,stachyose, and verbascose,215 are of particularinterest. Soybeans contain (by weight) approxi-mately 1% of raffmose and 2.5% of stachyose;215

the winged bean contains 1-2% of raffmose, 2-4% of stachyose, and 0.2-1% of verbascose.94

Ptaquiloside

Bracken fern(s) (Pteridium spp.) foundthroughout the world causes cancer of the uri-nary bladder in ruminants and is the only higherplant shown to cause cancer naturally in ani-mals.249 Enzootic hematuria, the clinical namegiven to the urinary bladder neoplasia of rumi-nants, tends to occur persistently in localizedbracken-infested regions. The major carcinogenof bracken is the mutagenic and clastogenicnorsesquiterpene glucoside, ptaquiloside, whichin laboratory animals has been shown to be carci-nogenic per a?.1 1U90'249 Bracken has further beenassociated with carcinoma of the upper digestivetract of cattle, where it is believed to transform thebovine papilloma (type 4) to a malign tumor. Afterhydrolysis of the glucoside, the genin is partlyconverted under alkaline conditions to adienone, which can then undergo further reac-tions to form adducts with DNA bases. A pre-liminary investigation of the alkylation patternsproduced has been presented.250 Bracken fern isacutely toxic to several farm animals such ashorses, cattle, and sheep, the syndromes being

different for the different animal species.82 Theadministration of pure ptaquiloside to a calf re-sulted in the same symptoms as known for thebracken intoxications of this species, thus dem-onstrating that the causative principle of cattlebracken poisoning is ptaquiloside.112

Saponins

A great number of food and feed plants con-tain saponins. Saponins may belong either to thegroup of pentacyclic triterpenoid saponins or tothe steroidal saponins. The latter include in abroad sense the steroidal alkaloid glycosides thatare found, for example, in potatoes. Althoughcertain saponins such as the medicinally usedquillaja saponin have been known for centuriesto damage mucous membranes,125 most saponinsare considered quite unproblematic when theyare administered orally. Saponin fractions fromcertain Yucca spp. have even been used as a feedadditive to promote growth of, for example, tur-keys.71 However, concerns have been raised re-cently that saponins in food or feed may promoteoral sensitisation to allergens through theirmembranolytic action in the gastrointestinaltract, resulting in enhanced uptake of the aller-gens.128 This concern is based on the fact that sa-ponins have been shown to act as oral adju-vants.59'132'158 Food plants that may containconsiderable amounts of saponins include theseeds ofQuinoa spp., fruits of Borassus flabellifer(palmyrah) and Balanites spp., and roots and sto-lons ofGlycyrrhizaeglabra (licorice) (Table 4-1).The palmyrah fruits are fermented to wine (palmwine) in Sri Lanka, whereas experimental solid-state tempeh fermentations have been describedfor quinoa. However, no information is availableconcerning the fate of the saponins in any of theseproducts.

Vicine and Convicine

Vicia faba (faba/fava bean), V. harbonensis,and V. sativa contain the two glycosides, vicineand convicine,219 which after hydrolysis in theintestine and uptake of the genins (divicine andisouramil) cause hemolytic anemia (favism) in

glucose-6-phosphate dehydrogenase-deficientindividuals.166'284 Together with condensedtannins, these two glycosides limit the use of theproteinaceous raw faba beans as feed for mono-gastric animals.167'276 Vicine and convicine havenot been detected in significant concentrationsin other plant species.

Cucurbitacins

Cucurbitacins were first characterized as thebitter compounds of cucumbers, marrows, andsquashes (Cucurbitaceae). The Cucurbitacins asa group are thought to be among the most bittersubstances known to man. Cucurbitacin B canbe detected in dilutions as low as 1 ppb, and theglycosides of cucurbitacin E at 10 ppb.175

Cucurbitacins make up a group of oxygenatedtetracyclic triterpenes, some of which occur asglycosides.103 Some Cucurbitacins are not onlybitter, but also toxic. Thus, the lethal dose for10% of a test group of mice (LD10 orally mice) ofcucurbitacin B is approximately 5 mg/kg b.w.103

This is quite strong toxicity, as seen from thefact that it is equal to the lowest dose used in theInternational Organization for Economic Coop-eration and Development Guidelines test foracute oral toxicity (Fixed Dose procedure,guideline no. 420).

Glucosinolates

In 1990, more than 100 glucosinolates werealready known.252 They occur in Capparales,Salvadorales, Violates, Euphorbiales, andTropaeolales within Violiflorae sensu Dahl-gren.57'58 Reasons for interest in glucosinolatesor glucosinolate-containing plants are the vari-ous antinutritional and toxic effects, the flavors,and the positive physiological effects associatedwith these constituents and their byproducts.23

Rape (Brassica napus, B. campestris, and B.juncea) is among the most important crop con-taining glucosinolates. Seeds of these speciescontain approximately 400 g of oil and approxi-mately 250 g of protein per kg. However, the useof rapeseed meals as a protein source in live-stock rations and human diets is limited because

of compounds associated with the protein frac-tions. These include phytic acid, phenolic com-pounds, and glucosinolates. Rapeseed that isbred to contain less than 2% erucic acid in its oiland less than 30 |ig/g of aliphatic glucosinolatesis termed "double low" or "canola." All pureglucosinolates tested in animal diets have causedantinutritional or toxic effects even when theywere in concentrations relevant to levels basedon double-low rapeseed as the protein source.243

RISKS ASSOCIATED WITH THEOCCURRENCE OF TOXICGLYCOSIDES IN DIFFERENTCOMMODITIES

Regarding toxins in food, the compounds thatcall for discussions in further detail are the cya-nogenic glycosides, but also the MAM glyco-sides, ptaquiloside, the saponins, the favismagents (vicine and convicine), and theglucosinolates.

Cyanogenic Glycosides

Although discussions concerning a toxicity ofintact cyanogenic glycosides may be found, theliterature at present concludes that known in-toxication syndromes, whether acute or chronic,are mainly due to HCN that is formed from thecompounds.235'236

A plant containing cyanogenic glycosidesmay or may not contain enzymes that catalyzetheir breakdown (i.e., hydrolases [(3-glycosi-dases] and cyanohydrin lyases). These are storedseparately from the glycosides.211 When a tissuecontaining both cyanogenic glycosides andthese enzymes is crushed, enzyme(s) andsubstrate(s) are brought together and hydrolysisand further lysis (i.e., cyanogenesis) starts. Thus,the intake of raw or processed cyanogenic mate-rial normally will mean an intake of a mixture ofthe genuine glycoside(s) and accompanying hy-drolysis products. Tissues that only contain theglycosides (and not the enzymes) will only giverise to exposure to the genuine glycoside(s).

Although cyanogenic glycosides may un-dergo acid hydrolysis,30'83 the conditions in the

stomach of a nonruminant, together with thevery short residence time, will let the main frac-tion pass to the intestine. In the intestine, the gly-cosides will be absorbed, as shown for linamarinin a number of animal species and in hu-mans,19'37'110'208 and for prunasin and amygdalinin different animal species,44'220'221'235 or it will behydrolyzed by microorganisms.28'44'210 In hu-mans, Carlson et a/.43 very recently found thatapproximately 25% of linamarin ingested in astiff porridge prepared from cassava flour wasabsorbed and excreted unchanged in the urine,whereas a little less than 50% was converted tocyanohydrin or cyanide and absorbed as such.The rest could not be accounted for. Most or allof the absorbed glycosides will be excreted inthe urine, as shown for both linamarin andamygdalin in animals and humans.7'37'110'173 HCNas well as cyanohydrins will give rise to cyanideexposure through absorption and nonenzymaticlyses of the cyanohydrins.

Whereas the absorbed glycosides will be ex-creted unchanged in the urine, the HCN will betotally or partly metabolized, the main metabo-lite being the goitrogenic compound, thiocyan-ate. The rate of this conversion will depend onthe nutritional status of the individual. Currentknowledge concerning the known biomarkersfor cyanide exposure (acute and long term) andtheir use in clinical and experimental toxicologywas reviewed by Rosling.231 The detoxificationprocesses (metabolization) and methods for theestimation of the sulphane sulphur pools availablefor this were reviewed by Westley.281 Acute hu-man intoxications have been described as a resultof the intake of cassava products and almonds,whereas sorghum and cyanogenic acacia leavesand pods have caused veterinary intoxications.

Acute Intoxication

Acute intoxications in humans caused by theintake of insufficiently processed cassava mealshave been reported from nearly all parts of the cas-sava consuming area, although it must be empha-sized that the published reports are very scarce inrelation to the extensive use of cassava as humanfood.5'76'78'172 The symptoms of acute intoxication

include vomiting, nausea, headache, dizziness,difficulty with vision, and collapse. 172

Chronic Intoxication Syndromes

Evidence has accumulated that cyanide expo-sure from the diet is a causative factor inkonzo,115'269 and may aggravate iodine defi-ciency disorders.60 The influence, if any, on thedevelopment of special types of diabetes re-mains a matter of discussion.3'263 Symptoms anddiagnosis of konzo have been described byRosling & Tylleskar.232

Based on the knowledge available concerningthe toxicity of cyanide and cyanogenic glyco-sides, the Joint World Health Organization(WHO)/Food and Agricultural Organization(FAO) Expert Commitee on Food Additives andContaminants (JECFA) tried to estimate a safelevel for the intake of cyanogenic glycosides byhumans. The committee concluded that "be-cause of lack of quantitative toxicological andepidemiological information, a safe level of in-take of cyanogenic glycosides could not be esti-mated." However, the committee also concludedthat "a level of up to 10 mg HCN/kg of productis not associated with acute toxicity."253(P332)

Thus, no authority has yet felt confident to setscientifically based safe levels for the intake ofone or more of the known cyanogenic glycosides(or their products of degradation), that is, levelsthat take the risk(s) for the development ofchronic intoxications into consideration. Inacknowledgement of this, the "InternationalWorkshop on Cassava Safety," held in Ibadan,Nigeria in 1994, concentrated on making recom-mendations concerning steps to be taken in re-search; in breeding programs; and in informa-tion to extension workers in the agricultural,food, and nutrition sectors.10

Long before humans knew the identity of cya-nide, they did know that bitter cassava is a goodstarch crop, but that it must be detoxified beforeconsumption.67'69 Today, we know that this isbecause of its content of the cyanogenic gluco-sides, linamarin and lotaustralin. Overviews ofthe occurrence of cyanogenic glycosides inplants used for human or animal consumptionare provided in Conn51 and Jones.129 Some of the

important species of plants have been subjectedto selection/breeding for a low total cyanogenicpotential (TCP). Examples of the constituentsand the TCP of economically important cropsare provided in the following sections, togetherwith some remarks concerning their importanceas food or feed commodities.

Phaseolus lunatus (Seeds) and Other Beans.Seeds from several species of legumes are usedfor human consumption, many of which containtoxic and antinutritional substances. Thus, seedsfrom, for example, P. lunatus, P. aureus,Cajanus cajan, Canavalia gladiata, and Vignaunguiculata have been examined due to con-cerns about the possibility for cyanide intoxica-tions.63'192 All of these species are known to becyanogenic in one or more tissues.242 P. lunatuscontains linamarin as its main cyanogenic con-stituent; the cyanogens have not been identifiedin the other species.242 Only P. lunatus has beensubjected to investigations concerning the varia-tion in the cyanogenic potential.21 However, sev-eral of the other species certainly may containtoxic amounts of cyanogens in the seeds.192

Prunus Species (Seeds). Peach, plum, cherry,apricot, and almond (family Amygdalaceaesensu Dahlgren ) are all drupes (stone fruits) ofgreat importance to man. Cyanogenic glycosidestypical for Amygdalaceae are phenylalanine de-rived.182 Thus, the ripe seeds of P. persica(peach), P. domestica (plum), P. avium/cerasus(cherry), P. dulcis (P. amygdalus) (almond), andP. armeniaca (apricot) all contain amygdalin asthe major cyanogenic constituent. The total cya-nogenic potential per gram dry weight of wholefruit rises during the early development, and therelative composition of cyanogens changes from100% prunasin in the beginning to nearly 100%of amygdalin in the ripe seed.90'170'189 Amygdalinand different Prunus seeds have, in spite of theirineffectivity, been commercially promoted foryears as medicines to treat different cancers.108

• Almond — This tree is very widely culti-vated around the Mediterranean. The nam-ing of the species and its varieties/cultivars

has changed through time.99 The tree comesin two varieties, var. dulcis and var. amara,of which var. amara contains high concen-trations of amygdalin in its ripe seeds (alsodenoted "bitter almonds").39'50'99 The seedsare used in confectionary and bakery.99

They contain approximately 50% of lipids,the oil being used in cosmetics and derma-tology.39'99 Bitter almonds (but also, e.g.,apricot seeds) are also used to produce anessential (volatile) oil called "oil of al-monds." This competes with synthetic ben-zaldehyde as a source of flavor.39 Only fewreferences exist concerning the content ofcyanogenic glycosides in almonds.50'90

Conn50 found bitter almond seeds to release290 mg of HCN/100 g of seed. Accordingto Sturm,260 commercial sweet almondsfrom California in general contain lessbitter seeds (approximately 1%) than the2-3% that is normally found in the Medi-terranean ones.

• Apricot — Apricots have considerable eco-nomic importance for several countriessuch as Italy, the production of which wasapproximately 200,000 tons in 1988.259

Different products are marketed from apri-cots, including fresh, dried, and cannedfruits; nectar; jam; and distilled liqueur.176

The number of varieties and hybrids ofapricots are numerous.174 Thus, Audergonet al. 14 tested more than 400 varieties aspart of a physicochemical characterizationprogram. Several marketed products ofapricots require destoning,55 leaving thestones as a byproduct from which oil can beextracted. The use of the seed/presscake is,however, restricted by the toxicity.266 De-pending on the variety and type of apricot,the apricot stone is relatively small, repre-senting 6-8% of the fruit weight, even if itcan sometimes reach 10%.176 To the bestknowledge of the present author, no inves-tigations have been published concerningthe variation in content of amygdalin inseeds of different cultivars. However, aspart of investigations concerning the mi-crobial degradation of cyanogens in such

seeds, Tuncel et al. 266'267 analyzed twobatches of bitter and one of sweet Turkishapricot seeds, obtained on the commercialmarket. The bitter ones were found to con-tain approximately 52 and 92 |imol/g d.w.,respectively; the sweet ones contained ap-proximately 2.5 |imol/g.

• Peach — Much of the same that has beensaid for apricot can be said for peach. Seedsfrom P. persica Batsch (peach) also containamygdalin as their major cyanogenic con-stituent.193 Kupchella & Syty141 analyzedthe total cyanogenic potential of the seedsfrom an undefined cultivar and found it tocorrespond to a content of amygdalin of ap-proximately 2.45% w/w.

Linum usitatissimum (Seeds = Linseed/Flax-seed). Flax is grown for two main purposes, fi-bers and seeds. Different cultivar s are used forthe two products. Whole seeds are used as alaxative due to the swelling seed coat polysac-charides.200 Both full-fat flaxseed flour and de-fatted meal from the oil extraction are on thecommercial market, the latter in two qualities,with 30% and 40% protein, respectively.198 Flaxis one of the major industrial oilseeds traded inworld markets. Global production for crop year1994-1995 was 2.44 million metric tons, withCanada contributing a major share. Flaxseed oilis used for a multitude of purposes, the oil beingpriced up to four times that of the whole seed.198

The extraction cake (linseed meal) is tradition-ally used for fodder purposes. Recently, researchinto the refinement of flax products hasaccelarated. Thus, two patents have been issuedfor the use of flaxseed polysaccharide (gum) forcosmetic and medical preparations,13'196 and anoptimization of protein extraction from defattedflaxseed meal has been presented.199

Until 1980, linamarin was thought to be themain cyanogen in linseed. However, looking forthe factor(s) in linseed meal responsible for itsprotective effect against selenium toxicity,Smith et a/.251 isolated two new cyanogenic gly-cosides (linustatin and neolinustatin). A laterTLC-based investigation concerning the con-centrations of different cyanogenic glycosides in

a linseed sample gave the following |umol/g:linustatin+neolinustatin 4.6, linamarin 0.46, andlotaustralin 0.36,32 pointing to linustatin andneolinustatin as the major cyanogenic constitu-ents. This was further confirmed by a high per-formance liquid chromatography (HPLC) analy-sis of 48 samples, which on the other hand onlyfound traces of linamarin and lotaustralin.240

However, a recent investigation showed quitesome variation between 10 cultivars. Two con-tained no linamarin, whereas in the cultivarVimy, 7.8% of the weight of the total cyano-genic glycosides were linamarin.201 This is closeto the findings of Brimer et al}2 for an unspeci-fied sample. Frehner et a!.90 analyzed both thecyanogenic potential and the relative cyanogencomposition during fruit development — one cul-tivar. As in Prunus seeds, the monoglucosidespredominated at anthesis, shifting towarddiglycosides during maturation. Rosling230

found the cyanogenic potential of a nonspecifiednumber of commercially sold linseed in Swedento range from 4 mmol/kg to 12 mmol/kg (112-336 mg kg-1 HCN). The acute lethal dose is lessthan 2 mmol in 24 hours in sick and malnour-ished patients.270

Manihot esculenta (Roots and Leaves). Thegenus Manihot (Euphorbiaceae) incorporatesmore than 200 species, all originating in tropicalAmerica, from where several have been spreadto other continents. Thus, M. esculenta Crantz(cassava) is today grown as a major source ofstarch in tropical Africa, India, Indochina, Indo-nesia, and Polynesia.184 As early as 1605,Clusius reported that cassava could be toxic toman. The two cyanogenic glucosides, linamarinand lotaustralin, are responsible for this.70'183 Thecyanogenic potential (CNp) of several cassavagermplasm collections has been investigated.Thus, Aalbersberg & Limalevu1 analyzed 28cultivars grown in Fiji, and found a variationfrom approximately 15 mg to 120 mg HCNequivalents/kg f.w. Dufour67'69 looked at 14 cul-tivars of the Tukanoan Indians in northwestAmazonia and found very high levels (310-561mg HCN eq./kg f.w.) in KU (toxic cultivars) and171 mg HCN eq./kg f.w. in the only Makasera

(nontoxic/safe cultivar) grown. The TukanoanIndians clearly expressed that they preferredtoxic varieties as the main staple (70% of calorieintake) component of their diet. In this connec-tion, it should be noted that the so-called "safe"(Makaserd) cultivar had a higher CNp than the100 ppm (f.w.) that was proposed as the upperlimit by Koch67'69 based on acute toxicity. Fi-nally, Bokanga24 examined 1,768 different cas-sava collections and found that the content of thecentral pith of the root varied from approxi-mately 1 mg to more than 530 mg HCN equiva-lent/ kg d.w. The peel surrounding the pith has amuch greater content, as have the leaves.24 Noacyanogenic cassava was found. While discuss-ing the cyanogenic potential in this precise way, itshould be born in mind that variations of up to100% may be recorded between roots of the sameplant.24 It has also been shown that age, agricul-tural practices,73 and environment may have astrong influence on its cyanogenic potential.24'26

The leaves of M esculenta also serve as foodand feed.25 The cyanogenic potential of leavesfrom the same plant is less variable than that ofthe roots,24 and is usually 5 to 20 times higher ona fresh weight basis.25 The high content in theleaves normally does not present a problem fortheir use in food, given the methods generallyused in their preparation.25 In contrast, the rootsof many cultivars, if not properly processed,have actually caused both acute and chronic in-toxications worldwide. However, it should beemphasized again that the cassava root (evenhighly cyanogenic types) is a very valuable andirreplaceable crop. To ensure its safe use in ev-ery community under all conditions, the effec-tiveness of the different processing techniques(under rural as well as industrialized conditions)needs to be verified and the knowledge spread.10

Sorghum Species (Leaves and Seeds). Seed-lings of S. bicolor (Poaceae) and other Sorghumspecies synthesize the cyanogenic glucosidedhurrin that is localized in the ariel shoots of theplant.101 Thus, three-day-old etiolated seedlingsof S. vulgare (i.e., the name for any cultivatedgrain sorghum) was found to contain up to 15(imoles/g.4 The content in mature leaves is much

lower. The concentration depends on species,subspecies, and race/cultivar, and is also influ-enced by ecological factors.65 Although most in-toxications are seen in cattle browsing a newlysprouted field, forage may not be totallysafe.65'282 Grain sorghums constitute an impor-tant part of human nutrition in several semi-aridareas of the world.61'88 Generally, the grains areconsidered completely safe for human consump-tion,88'136 although the digestibility and biologi-cal value are not always high as a result of theoccurrence of quite high concentrations ofphytate and polyphenolics in several cultivatedtypes.88'120 Especially in Sudan, sorghum is irre-placeable, being the traditional stable food.64 Al-though sorghum seeds in general are safe, ger-minated seeds are not. In certain Africancountries, germinated sorghum seeds are usedtraditionally for the production of malt,88 whichin turn is used for the brewing of alcoholic bev-erages64 and for the production of the bakedproducts called Hulu-mur.64 According toFAO,88 the traditional methods of preparation ofthese products remove the dhurrin effectively;however, it is stressed that the existence of theseproducts must not be seen as an indication ofsprouted sorghums being safe — they are not.88

MAM Glycosides

A metabolic fate and mechanism of toxicity,including the same alkylating end product aswith dimethylnitrosamine, has been proposedfor the MAM that is released from the MAMglycosides.209 Thus, cycasin has been shown tobe toxic to a number of animals, causing hepaticlesions and demyelination with axonal swellingin the spinal cord.22' 246 Cow's milk may be a vec-tor of transmission of plant toxins. Thus,Mickelsen et a/.168 showed that MAM can passinto the milk of lactating rats, causing tumors inthe offspring. The seeds of several Cycas spp.are traditionally eaten in Australia22 and on cer-tain islands.150'254 A special neurological syn-drome occurring on the island of Guam, andtermed Guam ALS-PDC, has been hypothesizedto be due to the intake of seeds of C.circinalis.150'254 In 1987, Spencer et at.254 pro-

posed that the causative factor of this syndromewas the neuroexcitotoxic amino acid p-N-methylamino-L-alanine (BMAA). However, anumber of subsequent investigations doubtedthis, as reviewed by Stone258 in an article on thegradual disappearance of this disease. Thus, itmay never be known whether the MAM glyco-sides could have a role in this disease, though itremains a possibility given the spinal cord le-sions reported in goats as a result of chronic in-take of cycasin.246

Ptaquiloside

The carcinogenicity of ptaquiloside demon-strated in feeding experiments with rats, mice,hamsters, guinea pigs, and cattle, among others,is alarming because the young shoots of brackenfern are highly regarded as a tasty dish in Ja-pan.111 Hence, this intake of bracken has beenlinked to high incidences of stomach cancer inJapan,111 and in Costa Rica among people whohave been exposed to milk that was produced inbracken-infested grasslands.6 The theory hasbeen supported by the finding of a high tumorincidence in rats and mice that were fed milkfrom cows that had been fed with dietarycomplements of bracken, and by the subsequentdemonstration of ptaquiloside in bovine milk.6

Saponins

Food and feed containing saponins includesoybean, guar, quinoa, balanites fruits, and oth-ers. Besides the membranolytic action of manysaponins, certain of these compounds exert spe-cial effects due to the structure of their agly-cone.286 Such effects include (1) lowering ofblood cholesterol;144 (2) reversible sodium reten-tion and potassium loss leading to hypertension,water retention, and electrolyte imbalance (e.g.,glycyrrhizinic acid found in licorice root, theroots and stolons from Glycyrrhiza glabra, andfor products to which licorice root extract, orglycyrrhizinic acid, has been added);100'133'239'256

and (3) crystal formation in the liver and biliarysystem, which may inhibit the excretion ofphylloerythrin (from chlorophyll degradation),

causing a subsequent photosensitation as seen in"Geeldikkop" (a Tribulus terristris intoxica-tion).131 A number of saponins are bitter. Theoccurrence of bitter saponins in palmyrah(Borassus flabellifer L) fruit pulp thus reducesthe use of juices based on this fruit.74 Likewise,seeds of Chenopodium spp. used for human con-sumption (C. quinoa [quinoa], C. pallidicaule[canihua], and C. berlandieri ssp. nuttaliae[Safford] Wilson and Heiser [huauzontle]) containbitter saponins,95'107'226 most of which are concen-trated in the outer layers of the grain.40'41'223

Vicine and Convicine

Favism is characterized by anemia, jaundice,and hemoglobinuria, and may develop in sub-jects with glucose-6-phosphate dehydrogenase(G6PD) deficiency as a consequence of fababean intake. Favism has also been reported inbreast-fed infants whose mothers had eaten fababeans, and in newborn infants.54 More than 300variants of G6PD are known.273 In addition, anassociation between the genotype OfACP1 (hu-man red cell acid phosphatase) and favism hasbeen shown, and a possible biochemical mecha-nism has been proposed.27 Most cases of G6PDdeficiency described in the past were from Italyand other countries around the Mediterranean,that is, patients with the common MediterraneanB-form of G6PD, rather than the common AfricanA (-) form.273 However, recent investigations haveshown that subjects with variants that result in arelatively mild G6PD deficiency may also developfavism.93'181 Preventive measures and treatmentshave been described elsewhere.102'162'188

Glucosinolates

The most prominent toxic manifestation ofglucosinolates in humans is the occurrence ofgoiter.214 In animal experiments, this and othereffects were generally more pronounced whenmyrosinases were included in the diet.252 The ef-fects seen were related to differences in the sidechains and to chirality.252 The fact that there areseveral mechanisms behind the toxic andantinutritional effects has also been very re-

cently stressed by the results of the most detailedstudies on the degradation products of variousglucosinolates.23 These authors presented anoverview of the different degradation productsformed from glucosinolates, which also include,for example, oligomers. From the degradation ofglucobrassicin (an indole glucosinolate),indolyl-3-methanol is formed in considerableamounts, but it disappears very quickly, givingrise to, among others, appreciable amounts ofthiocyanate ion. No organic isothiocyanates andthiocyanates are formed. In contrast, the degra-dation of various aliphatic glucosinolates resultsin the formation of nitriles as well asisothiocyanates and thiocyanates.23 Toxic effectsof glucosinolates in B. oleracea have been re-viewed by Stoewsand257 and those of crambe(Crambe abyssinicd) meal fed to broiler chicksby Kloss et al. 137 The mechanism behind the ob-served decrease in cancer risk for people on dietswith a high content of cruciferous vegetables hasbeen investigated by Wallig et a!.214

VARIATION IN TOXINCONCENTRATION AMONGVARIETIES AND CULTIVARS: THEINFLUENCE OF TRADITIONALDOMESTICATION AND MODERNBREEDING

Several toxic glycosides (including varioussaponins and cyanogenic glycosides, etc.) areknown to be bitter tasting in addition to toxic.Hence, the term "bitter," as opposed to "sweet,"has been used traditionally to designate naturallyoccurring or selected groups within a plant spe-cies that contain high amounts of the toxic (andbitter) substance. Depending on the view of thebotanical author, the groups in question may bedivided on the level of variety, form, or cultivar.Examples of plant species for which the divisionbitter/sweet has been used are P. dulcis and otherPrunus spp. (containing amygdalin), as well asM. esculenta (cassava, containing linamarin),and in quinoa.139 In most such cases, a correla-tion between the toxicity (content of glycoside)and the degree of bitterness of the plant part hasbeen established. However, it is only seldom that

a proper investigation concerning the degree towhich this correlation holds has been performed.Thus, a positive correlation, but with exceptions,was found in a number of smaller studies on cas-sava roots.247 Hence, King & Bradbury135 tookup the challenge of investigating in more detailthe bitter-tasting substances in cassava paren-chyma and cortex. Linamarin was found to bethe sole contributer to bitterness present in theparenchyma; a new structure (isopropyl-p-D-apiofuranosyl-( 1— 6)-(3-D-glucopyranoside) con-tributing in the cortex of some cultivars. This isin agreement with a very recent study fromMalawi, 234 which compared the content of cya-nogenic glucosides in the cortex of 492 cassavaroots with their taste as estimated by a tastepanel. The correlation had an r2 = 0.96 whenlooking at the cultivar level.

It is well documented, at least for a number ofcyanogenic plant species, that the concentrationof both the glycosides and the enzymes degrad-ing them can show a discrete variation (poly-morphism) as well as a continous one. The poly-morphism is genetically based, whereas thecontinous variations observed may be both ge-netically and environmentally influenced.24'26'116-119'130'159'186The genetic polymorphism (discretevariation, chemical races) with respect to the oc-currence of both cyanogenic constituents andhydrolytic enzymes makes it difficult to definewhat is meant by a "cyanogenic species." Fur-thermore, it should be noted that the cyanohy-drin lyase, which cleaves the cyanohydrinsformed after the hydrolysis of the glycoside(s),may be expressed in certain organs and not inothers. Thus, White et a/.283 recently showed thatthis enzyme, although present in the leaves, isnot expressed in the roots of cassava. This obser-vation explains why very high intermediate con-centrations of cyanohydrins are formed during theprocessing of cassava roots. The environmentalinfluences mentioned above may furthermoremean that certain plants will be found positive atsome times of the year and negative at others.

Increased use of more highly cyanogenic cul-tivars of cassava among small farmers has beenreported from several places. Thus, Dufour re-ported on a clear preference for Kii (toxic variet-

ies) for most purposes by the Tukanoan Indi-ans,67'69 whereas Aalbersberg & Limalevu1

stated that planting of the toxic (bitter) cultivarsincreased in New Guinea. Also, Onabolu et al 197

found that the three most commonly grown cul-tivars in Ososa (a semi-urban farming commu-nity approximately 80 km east of Lagos, Nige-ria), where cassava has been the main staple fordecades, were all stated to be poisonous and toneed processing. However, this was not re-garded as a disadvantage. Farmers' reasons forpreferentially growing cassava cultivars provid-ing bitter roots were studied in Malawi.48 Inmany traditional agricultural communities, thefarmers (often the women) judge the "safeness"of the roots by chewing a small piece. Accordingto Dufour, 67'69 the Tukanoan Indians appear to beable to distinguish accurately less from more poi-sonous cultivars by the taste. Very recent studiesfrom Malawi prove such a procedure to exist, andto be very effective (Chiwona-Karltun, personalcommunication, October 2000).

Several of the species within the familyCucurbitaceae, which are used as human food,naturally contain cucurbitacins in amounts thatare unacceptable to the market. However, in-tense domestication and breeding have resultedin cultivars low in bitter compounds.126'127

Breeding programs for curcurbits are constantlyaware of the bitterness.84

Great variations (0-13000 M£/g) may also befound in the content of ptaquiloside in brackenfern as a result of both ecological and geneticvariation, a tendency for higher contents beingreported when originating in relatively colderclimates.249 In addition, P. esculentum containsthe cyanogenic glucoside, prunasin, the concen-tration of which similarly has been related to cli-matic conditions.153

Also, quinoa cultivars vary concerning thequantitative content of saponins, and the tradi-tion has, as for other crops, been working withso-called sweet and bitter varieties.139

For V. faba, it should be mentioned that, al-though Due et al 66 gave the first report of a genethat codes for nearly a zero content of vicine andconvicine, present-day cultivars contain ap-proximately 7 and 2.5 mg g-1 respectively.276

REMOVAL OF TOXINS THROUGHPROCESSING

Traditional Household versus ModernIndustrial Processing

When discussing the removal of toxic andantinutritional constituents, one must distin-guish between traditional household processingand industrial processing. The two proceduresmay use different starting materials and willhave different means of analyzing these and dif-ferent methods available for processing. The pri-orities may indeed be very different when choos-ing between slow versus fast processingmethods, and between processes that require lowinput of water and/or energy as compared tomethods requiring a high input. The generaltrend of a greater number of traditional food fer-mentation processes being industrialized hasbeen discussed recently by Rombouts & Nout.228

A number of industrial processing methods orlaboratory methods meant for industrial devel-opment were investigated quite early on for soy-bean (e.g., removal of oligosaccharides and pro-teinase inhibitors)47'56'91'109'224'245'248'261'264 and forcruciferous plants (glucosinolates).20'72'85'151'255

Later, linseed (flax, cyanogenic glyco-sides),149'156'157'275 jojoba meal (organic nitrileglycosides),2 citrus juices (limonoids andnaringin),105'145'212'229'244 cotton seeds (gossy-pol),179 and quinoa seeds (saponins)95'226 werefocused on. For commodities such as cassavaroots and leaves, lima beans, cycas seeds,bracken leaves, and yam tubers (alkaloids aswell as terpenes), most methods described andinvestigated scientifically are actually house-hold processing methods. A look at the process-ing of cassava roots will help illustrate the im-portant characteristics for each of the twosectors.

Household Processing (Cassava Roots)

Both in South America and in sub-SaharanAfrica, sweet and bitter roots are generally re-garded as two well-known different crops, andmost traditional methods of processing that havebeen studied have proven very effective in re-

Next Page

ies) for most purposes by the Tukanoan Indi-ans,67'69 whereas Aalbersberg & Limalevu1

stated that planting of the toxic (bitter) cultivarsincreased in New Guinea. Also, Onabolu et al 197

found that the three most commonly grown cul-tivars in Ososa (a semi-urban farming commu-nity approximately 80 km east of Lagos, Nige-ria), where cassava has been the main staple fordecades, were all stated to be poisonous and toneed processing. However, this was not re-garded as a disadvantage. Farmers' reasons forpreferentially growing cassava cultivars provid-ing bitter roots were studied in Malawi.48 Inmany traditional agricultural communities, thefarmers (often the women) judge the "safeness"of the roots by chewing a small piece. Accordingto Dufour, 67'69 the Tukanoan Indians appear to beable to distinguish accurately less from more poi-sonous cultivars by the taste. Very recent studiesfrom Malawi prove such a procedure to exist, andto be very effective (Chiwona-Karltun, personalcommunication, October 2000).

Several of the species within the familyCucurbitaceae, which are used as human food,naturally contain cucurbitacins in amounts thatare unacceptable to the market. However, in-tense domestication and breeding have resultedin cultivars low in bitter compounds.126'127

Breeding programs for curcurbits are constantlyaware of the bitterness.84

Great variations (0-13000 M£/g) may also befound in the content of ptaquiloside in brackenfern as a result of both ecological and geneticvariation, a tendency for higher contents beingreported when originating in relatively colderclimates.249 In addition, P. esculentum containsthe cyanogenic glucoside, prunasin, the concen-tration of which similarly has been related to cli-matic conditions.153

Also, quinoa cultivars vary concerning thequantitative content of saponins, and the tradi-tion has, as for other crops, been working withso-called sweet and bitter varieties.139

For V. faba, it should be mentioned that, al-though Due et al 66 gave the first report of a genethat codes for nearly a zero content of vicine andconvicine, present-day cultivars contain ap-proximately 7 and 2.5 mg g-1 respectively.276

REMOVAL OF TOXINS THROUGHPROCESSING

Traditional Household versus ModernIndustrial Processing

When discussing the removal of toxic andantinutritional constituents, one must distin-guish between traditional household processingand industrial processing. The two proceduresmay use different starting materials and willhave different means of analyzing these and dif-ferent methods available for processing. The pri-orities may indeed be very different when choos-ing between slow versus fast processingmethods, and between processes that require lowinput of water and/or energy as compared tomethods requiring a high input. The generaltrend of a greater number of traditional food fer-mentation processes being industrialized hasbeen discussed recently by Rombouts & Nout.228

A number of industrial processing methods orlaboratory methods meant for industrial devel-opment were investigated quite early on for soy-bean (e.g., removal of oligosaccharides and pro-teinase inhibitors)47'56'91'109'224'245'248'261'264 and forcruciferous plants (glucosinolates).20'72'85'151'255

Later, linseed (flax, cyanogenic glyco-sides),149'156'157'275 jojoba meal (organic nitrileglycosides),2 citrus juices (limonoids andnaringin),105'145'212'229'244 cotton seeds (gossy-pol),179 and quinoa seeds (saponins)95'226 werefocused on. For commodities such as cassavaroots and leaves, lima beans, cycas seeds,bracken leaves, and yam tubers (alkaloids aswell as terpenes), most methods described andinvestigated scientifically are actually house-hold processing methods. A look at the process-ing of cassava roots will help illustrate the im-portant characteristics for each of the twosectors.

Household Processing (Cassava Roots)

Both in South America and in sub-SaharanAfrica, sweet and bitter roots are generally re-garded as two well-known different crops, andmost traditional methods of processing that havebeen studied have proven very effective in re-

Previous Page

moving the cyanogens.17'68'197 Likewise, studiesin Mozambique,37'169 Tanzania,172 and Zaire18

have demonstated that problems with intoxica-tions, whether accute or chronic, are seen onlywhen shortcuts are made in the traditional meth-ods of processing. Shortcuts may be introduceddue to shortage of food or to a wish to produceproducts for the commercial market faster.

Industrial Processing (Cassava Roots)

Although industrial processing of cassavaroots for starch has existed for decades in Bra-zil,45 cassava processing industries are cominginto existence only gradually in Africa. Theseinclude small-scale producers of flours, biscuits,and fermented products such as gari,16'237 as wellas bigger industries making the same products orproducts such as dry snacks. If industrial pro-cessing starts from bitter cultivars, precautionsmust be taken to ensure that methods are fol-lowed that give safe products. Often, traditionalhousehold methods will be scaled up becausethey are well known and give products that arewell accepted by the local population. However,because these methods generally require pro-longed processing,15'202 it may be tempting tomake shortcuts in order to increase productionand/or reduce the requirements for storage/pro-cessing capacity, water, or energy. In such cases,it will be crucial to investigate the new processand the products.

Unit Operations that May Influence theConcentration or the Identity of Toxins

A reduction in the level of a toxin or anantinutritional constituent in a product may beachieved in different ways. Analyzing the reduc-tions reported in a number of studies on one ormore of the commodities such as soy-bean,42'47'56'151'233 rapeseed,20'72'85'255 yam tu-bers,123'278 jojoba meal,2 Cycas seeds,22 citrusjuices,75'105'145 quinoa seeds,95 faba beans,12'106'167'287

cassava roots and leaves,8'9'11'15'77'142'143'202 flax-seeds,156'157'275 lima beans and other beans,124'192'218

Prunus seeds,185'238'266'267 and Sorghum greenparts,282 one can conclude that substances may(1) leach during soaking (steeping/retting),

(2) be deliberately extracted using aqueous ornonaqueous solvents or two-phase systems, or(3) be degraded. The rate and extent of leaching/extraction will depend on the degree of commi-nution (particle size) or, for whole seeds orroots, on whether dehulling or peeling has takenplace. Tissue softening/maceration and cell dis-ruption by the action of growing microorgan-isms or added enzymes will also play a role.Degradation may be the result of a chemical re-action catalyzed by endogenous and/or addedenzymes. It may also be the result of a chemicalor heat treatment. Finally, the compound may bedegraded by microorganisms during a fermenta-tion. In all cases, the degradation will normallydepend on parameters such as moisture content,temperature, and pH. In the case of enzyme-catalyzed reactions and reactions with addedchemicals (e.g., oxidations), the substrate (thetoxin) must be released so as to come into con-tact. The method of grating, how effective cellrupture is, and the resulting particle size willhave an influence on the end result.266 Contactmay be facilitated further by the action of micro-organisms or of added pectinolytic and cellu-lolytic enzymes. As mentioned previously, thelevel of endogenous enzyme(s) present mayvary from cultivar to cultivar.

Again, one can learn from the production of acassava product, namely, the fermented product,gari. Early investigations into the hydrolysis oflinamarin during the production of gari reportedthat hydrolysis took days,121'134'191 whereasVasconcelos et a/.271 found that 95% of initiallinamarin was hydrolyzed three hours after grat-ing the roots. A fast degradation was also ob-served by Giraud et <2/.,97(p81) which led thisgroup to conclude: "The observed differencesmay be explained by the use of non-traditionalmeans for the preparation of gari, particularlyduring the grating of the cassava roots or asMkpong et al (1990) [see ref. 171] noticed, bythe utilisation of cassava varieties showing dif-ferent levels of endogenous linamarase." Both ofthese variables have already been touched on;the example thus further underlining the impor-tance of chemical control whenever a change ismade, either in starting materials, processing

equipment, or the process. However, there ismore in this quotation that calls for a discussion,specifically, the words "non-traditional meansfor the preparation." From this, one could get theimpression that because the two most recent pa-pers find a "fast" degradation, this is most prob-ably what one will get using "traditional equip-ments/methods for the grating." But, what are"traditional equipments/methods"? Obviously, aperson from Nigeria, the world's largest cassavaproducer89 (where the crop is mainly producedfor human consumption by subsistence farmers)and the home country of gari, would think of theequipment/methods used in his or her region as"traditional." Hence, it is interesting to note thata very recent investigation where local femaleNigerian processors prepared gari using theirown utensils could report a period of hydrolysisof approximately four days. A control study atthe research station International Institute ofTropical Agriculture (UT A) in Ibadan, Nigeria,using the same method gave the same result,even though two different cultivars were used inthe two examples.197

Degradation Products

The degradation of a toxic constituent maylead to one or more nontoxic reaction products;however, it may also lead to the formation ofnew toxins. For glycosides, the most commondegradation processes will be those that startwith the hydrolysis of the glycosidic bond(s)giving the aglycone and one or more saccha-rides. However, other routes of degradation havebeen described. Thus, the cyanogenic glucoside,linamarin, is metabolized to the glycoside of thecorresponding amide by a bacterium,Brevibacterium sp. strain R312,147'148 whereasVerbiscar et al212 reported that a number of strainsof Lactobacillus acidophilus and L. bulgaricusseem to degrade the toxic (noncyanogenic) cyano-glycosides in jojoba meal by acting on thecyanogroup. Returning to the route starting withthe hydrolysis of the glycosidic bond(s), oneshould still be aware that several different possi-

bilities often exist, depending on the glycosideand the enzyme/microorganism in question.Again, the degradation of cyanogenic glycosideswill be used as an example.

A cyanogenic p-bis-glycoside such as amygda-lin or vicianin may in principle be hydrolyzedthrough a sequential (two-step) mechanism or asimultanous (one-step) mechanism (Figure 4-3).Both systems have been found in cyanogenicplants.114 For degrading microoganisms, however,Brimer and coworkers34'36 demonstrated that mostseem to cleave such glycosides in two steps, al-though important differences in the overall pro-cess have been observed. Thus, a considerableconcentration of the intermediate productprunasin, another (toxic) cyanogenic glycoside,is formed during the hydrolysis of amygdalin by ayeast such as En domycopsis fibuliger and by fila-mentous fungi such as Mucor circinelloides andPenicillium aurantiogriseum. In contrast, onlynegligible concentrations were detected using astrain of L plantarum.149 In all four cases, the endproduct was the cyanohydrin mandelonitrile, atoxin in itself, and in general more toxic than theparent glycoside270 because the latter may be ab-sorbed and excreted unchanged.37

Depending on the product/process in whichthe cyanohydrin is formed, it may have differentfates. The commodity itself may contain a cy-anohydrin lyase (a-hydroxynitrilelyase) thatcatalyzes the cleavage to form HCN and anoxocompound (Figure 4-2). This is true for cas-sava leaves, almond, and linseed,165 but appar-ently not for cassava root, where the enzymeseems not to be expressed.283 Also, if the startingmaterial has been heated (e.g., an oil extractioncake), this process may not happen because en-zymes may have been inactivated. If the result-ing pH is higher than around six, the cyanohy-drin will disintegrate spontaneously to form thesame two products. Exogenous enzymes ormicrooganisms can also metabolize the cyano-hydrin;122'138 however, described examples onthe latter using food-acceptable processes arefew. In an alkaline medium, any cyanide that isformed will remain present if it is not extracted.

Figure 4-3 Different mechanisms for the enzymatic degradation (hydrolysis) of a cyanogenic glycoside (con-taining a disaccharide moiety) to a cyanohydrin, which in turn may release HCN either spontaneously or byenzyme catalysis, (a) Cleavage of a glycoside by a bis-glycosidase (simultaneous mechanism, e.g., Vicianin in aDavallia species), (b) Cleavage by the concerted action of two hydrolytic reactions (sequential mechanism, e.g.,the hydrolysis of the bis-glucoside amygdalin by glucosidase A and B in almond). Note: GIu = Glucose Moiety,Glue. = Glucose, GIy = a monosaccharide moiety (i.e., glucose or another).

In moist acid media, the cyanohydrin, if not fur-ther metabolized, will prove quite stable.87'97

However, it may be driven out by drying or byheating/cooking.197'266 Looking at the conse-quences of these examples, the following shouldbe noted.

1 . Degradation of a toxin may result in theformation of another. This means:a. When new processes/toxic raw materi-

als are taken into a production, thewhole degradation process should beclarified or the end product should betested toxicologically.

b. If a process that results in the formationof other toxins is used, the chemicalanalysis used for end-product controlmust quantify all possible toxins (start-ing, intermediate, and end products).

2. New toxins formed may be removed by thesame type of process(es) that led to degrada-tion of the parent compound(s) or may needother measures for their removal.

Microorganisms and Enzymes that CanFacilitate the Removal of Toxins throughDegradation

Recently, Reddy & Pierson222 published ashort review on the reduction of antinutritionaland toxic compounds in plant foods by fermen-tation. Chemical entities discussed includedphytates, tannins, cyanogenic glycosides, ox-alates, saponins, lectins, and inhibitors of en-zymes such as oc-amylase, trypsin, and chymot-rypsin. Phytates may be hydrolyzed byendogenous as well as microbial phytases. Thereduction of phytates during breadmaking andtempeh fermentations were discussed in particu-lar. Most of the parameters pointed to earlier inthis text are shown to be of importance, depend-ing on the product and raw materials. Among theglycosides, only cyanogens are discussed andonly briefly. Hence, more details concerningcurrent knowledge about enzymes and microor-ganisms that can degrade toxic glycosides willbe provided next. A review of all processes and

bis-glycosidase

all organisms shown to be able to degrade one oranother toxin is beyond the scope of this article.Thus, reference will only be made to a limitednumber of review articles and original papers.

Microorganisms

A number of microorganisms isolated fromfermenting food and feed products have been in-vestigated for their ability to degrade toxins, es-pecially glycosides. Likewise, a number ofscreenings have been performed on other spe-cies/strains from different collections of micro-organisms. The glycosides used as substrateshave mainly been the cyanogenic glycosides,amygdalin,33'38 linamarin,35'285 and linustatin;149

the jojoba cyanoglucosides;272 and vicine/convicine.164 The resulting overall picture is thathigh p-glycosidase activities are most oftenfound within filamentous fungi, followed byyeasts and, finally, bacteria. Among the bacteria,lactic acid bacteria such as L plantarum tend toshow relatively high activities. The level of ac-tivity is strongly strain dependent. Compared toplant enzymes,114 most crude as well as purifiedmicrobial p-glycosidases seem to possess quite abroad substrate specificity,149'207 a factor thatmay be important to remember and analyze infurther detail, particularly when screening usingan artificial substrate.149 In addition to the abilityto degrade the glycosides, it is also crucial thatthe organism grows well on the vegetable sub-strate.185 Degradation of toxic glycosides bymeans other than hydrolysis of the glycosidiclink has only been described in a few cases: Thecase of the bacteria Brevibacterium sp., andstrains of L acidophilus and L. bulgaricus, havealready been mentioned.

Whether the vegetable toxin itself can inhibitthe growth of microorganisms has been ad-dressed in only a few studies. The study byBrabban & Edwards29 on the effect ofglucosinolates on microbial growth showed anantimicrobial effect of both the parent com-pounds and the aglycones formed as a result ofmicrobial hydrolysis of the glucosinolates. Thepaper is stongly recommended for its detailedand illustrative discussions.

Enzymes

Studies have been published regarding the useof added plant enzymes for the hydrolysis oftoxic glycosides.12'268'282 In the case of the addi-tion of (3-glycosidase from almond (as powderedalmond) to faba bean dishes, quite effectivetreatments have been demonstrated at relativelylow cost.12 However, most other studies haveshown a need for such high amounts of partlypurified enzyme(s) that it seems unfeasible froma cost-benefit point of view. Crude microbialenzymes with known glycosidase activities frommicroorganisms have also been investigated.Such studies have demonstrated effective accel-eration of detoxification processes in cassava207

or hydrolysis of pure glycosides from fababean.164 However, most such crude enzymeshave been derived from filamentous fungi, andthe relative role of the glycosidase(s) comparedto other activities present, such as cellulase andpectinases, is still to be clarified. A significantrole for cellulase in particular, in both the soften-ing of tissue and the acceleration of natural toxindegradation, has been demonstated for linamarindegradation in cassava root.8'9'11'77'79-81

CONCLUSION

A number of food plant species contain toxicor antinutritional substances. In cases where cul-tivars/varieties are grown that have a contentthat is too high for consumption, the level mustbe reduced through processing. A number ofsuch methods of processing have been devel-oped succesfully for a wide variety of cropsthroughout the world. The reduction of glyco-sides is especially well documented in literature,safe products being produced by the fermenta-tion of roots from even highly toxic cultivars ofcassava. However, incidents of both acute andchronic intoxications have been described, but ingeneral only when the usual methods of fermen-tation were not adhered to.

Several vegetables studied show great varia-tion among varieties/cultivars with respect to thecontent of both toxins and endogenous enzymes

that can cause changes in the level or nature ofthe toxins during processing. Likewise, the ac-tivity of enzymes such as (3-glycosidases,pectinases, and cellulases are strongly strain de-pendent when looking at the fermenting micro-organisms. Although glycosidases may contrib-ute directly to the degradation of glycosidictoxins, the tissue and cell wall degrading en-zymes may facilitate the detoxification process,releasing both the toxins and endogenous plantenzymes.

REFERENCES

1 . Aalbersberg, W. G. L. & Limalevu, L. (1991). Cyanidecontent in fresh and processed Fijian cassava (Manihotesculentd) cultivars. Trop Sd 31, 249-256.

2. Abbott, T. P., Nakamura, L. K., Buchholz, G., Wolf,W. J., Palmer, D. M., Gasdorf, H. J., Nelsen, T. C. &Kleiman, R. (1991). Processes for making animal feedand protein isolates from jojoba meal. J Agric FoodChem 39, 1488-1493.

3. Akanji, A. O. (1994). Cassava intake and risk of diabe-tes in humans. Acta Hortic 375, 349-359.

4. Akazawa, T., Miljanaich, P. & Conn, E. E. (1960).Studies on cyanogenic glycoside of Sorghum vulgare.Plant Physiol 35, 535-538.

5. Akintonwa, A., Tunwashe, O. & Onifade, A. (1994).Fatal and non-fatal acute poisoning attributed to cas-sava-based meal. Acta Hortic 375, 285-288.

6. Alonso-Amelot, M. E., Perez-Mena, M., Calcagno, M.P. & Jaimes-Espinoza, R. (1992). Quantitation ofPterosins A and B, and Ptaquiloside, the main carcino-gen of Pteridium aquilinum (L. Kuhn), by high pres-sure liquid chromatography. Phytochem Anal 3, 160—164.

7. Ames, M. M., Kovach, J. S. & Flora, K. P. (1978). Ini-tial pharmacologic studies of amygdalin (laetrile) inman. Res Commun Chem Pathol Pharmacol 22, 175-185.

8. Amoa-Awua, W. K. A., Appoh, F. E. & Jakobsen, M.(1996). Lactic acid fermentation of cassava dough intoagbelima. IntJFoodMicrobiom, 87-98.

9. Amoa-Awua, W. K. A., Frisvad, J. C., Sefa-Dedeh, S.& Jakobsen, M. (1997). The contribution of mouldsand yeasts to the fermentation of "Agbelima" cassavadough. J Appl Microbiol 83, 288-296.

10. Summary and recommendations: proceedings of theInternational Workshop on Cassava Safety, Ibadan,Nigeria, March 1-4, 1994. (1994). Acta Hortic 375,11-19.

In conclusion, the safety of the end productdepends on a number of factors such as the (1)overall process scheme, (2) overall time of pro-cessing, (3) equipment used (e.g., type of grind-ers), (4) variety/cultivar of vegetable commodityused, and (5) identity (genetic stability) of themicroorganisms used in fermentations. Anychange in any of these factors, such as when pro-duction is industrialized, should trigger a checkof the safety of the end product through chemi-cal analysis or by other means.

11. Amoa-Awua, W. K. A. & Jakobsen, M. (1995). Therole of Bacillus species in the fermentation of cassava.JApplBacteriol 79, 250-256.

12. Arbid, M. S. S. & Marquardt, R. R. (1985). Hydrolysisof the toxic constituents (vicine and convicine) in Fababean (Viciafaba L.) food preparations following treat-ments with p-glucosidase. J Sd Food Agric 36, 839-846.

13. Attstrom, R., Glantz, P. O., Hakansson, H. & Larsson,K. (1993). U.S. Patent 5,260,282.

14. Audergon, J. M., Duffillol, J. M., Souty, M., Breuils, L.& Reich, M. (1991). Biochemical and physicochemi-cal caracterisation of 400 apricot varieties: conse-quences in the apricot selection and improvement pro-cess. Acta Hortic 293, 1 1 1-1 19.

15. Balagopalan, C., Padmaja, G., Nanda, G. & Moorthy,S. N. (1988). Cassava in Food, Feed and Industry.Boca Raton, FL: CRC Press.

16. Banalya, J. (1998). Ghana hi-tech to process cassava.Post Harvest Systems (IITA) 3, 12.

17. Banea, M., J. -P., Nkiabungu, B., Tylleskar, T. &Rosling, H. (1998). High cassava consumption withoutcyanide exposure in Kinshase, in former Zaire. EcolFood Nutr 37, 363-377.

18. Banea, M., Poulter, N. & Rosling, H. (1992). Shortcutsin cassava processing and risk of dietary cyanide expo-sure in Zaire. Food Nutr Bull 14, 137-143.

19. Barrett, M. D. P., Hill, D. C., Alexander, J. C. &Zitnak, A. (1977). Fate of orally dosed linamarin inrats. Can J Physiol Pharmacol 55, 134-136.

20. Bau, H. -M., Villaume, C., Lin, C. -F., Evrard, J.,Quemener, B., Nicolas, J. -P. & Mejean, L. (1994). Ef-fect of a solid-state fermentation using Rhizopusoligosporus sp.T-3 on elimination of antinutritionalsubstances and modification of biochemical constitu-ents of defatted rapeseed meal. J Sd Food Agric 65,315-322.

21. Baudoin, J. P., Barthelemy, J. P. & Ndungo, V. (1991).Cyanide production of the lima bean, Phaseoluslunatus L. genetic variability in the primary and sec-ondary gene pools and in some intraspecific hybridepopulations. BullRech Agron Gembloux 26, 367-388.

22. Beck, W. (1992). Aboriginal preparation of Cycasseeds in Australia. Econ Bot 46, 133-147.

23. Bjergegaard, C., Li, P. W., Moeller, P., Otte, J. &Soerensen, H. (1994). Glucosinolates and their transfor-mation products — compounds with a broad biologicalactivity. In Bioactive Substances in Food of Plant Ori-gin. Vol. 1 , Low Molecular Weight N-Containing Com-pounds, Phenolic Compounds and Antioxidants, En-zyme Inhibitors, pp. 1-15. Edited by H. Kozlowska, J.Fornal & Z. Zdunczyk. Olztyn, Poland: CentrumAgrotechnologii i Weterynarii - Polska Akademia Nauk.

24. Bokanga, M. (1994). Distribution of cyanogenic poten-tial in cassava germplasm. Acta Hortic 375, 1 17-123.

25. Bokanga, M. (1994). Processing of cassava leaves forhuman consumption. Acta Hortic 375, 203-207.

26. Bokanga, M., Ekanayake, I. J., Dixon, A. G. O. &Porto, M. C. M. (1994). Genotype-environment inter-actions for cyanogenic potential in cassava. ActaHortic 375, 131-139.

27. Bottini, E., Bottini, F. G., Borgiani, P. & Businco, L.(1997). Association between ACPi and Favism: a pos-sible biochemical mechanism. Blood 89, 2613-2615.

28. Bourdoux, P., Mafuta, M., Hanson, A. & Ermans, A.M. (1982). Cassava toxicity: the role of linamarin. InNutritional Factors Involved in the Goitrogenic Actionof Cassava, pp. 15-27. Edited by F. Delange, F. B.Iteke & A. M. Ehrmans. Ottawa, Canada: IDRC.

29. Brabban, A. D. & Edwards, C. (1995). The effect ofglucosinolates and their hydrolysis on microbialgrowth. J Appl Bacteriol 79, 171-177.

30. Bradbury, J. H., Egan, S. V. & Lynch, M. J. (1991).Analysis of cyanide in cassava using acid hydrolysis ofcyanogenic glucosides. J Sd Food Agric 55, 277-290.

31. Brimer, L. (1998). Determination of cyanide and cya-nogenic compounds in biological systems. In CyanideCompounds in Biology; CIBA Foundation SymposiumNo. 140, pp. 177-200. Edited by David Evered and Sa-rah Harnett. Chichester, England: John Wiley & Sons.

32. Brimer, L., Broegger Christensen, S., Moelgaard, P. &Nartey, F. (1983). Determination of cyanogenic com-pounds by thin-layer chromatography. J Agric FoodChem 3 1,789-793.

33. Brimer, L., Cicalini, A. R., Federici, F. & Petruccioli,M. (1994). Production of p-glycosidases (linamaraseand amygdalase) and pectolytic enzymes by Penicil-lium spp. World J Microbiol Biotechnol 10, 203-206.

34. Brimer, L., Cicalini, A. R., Federici, F. & Petruccioli, M.(1998). Amygdalin degradation by Mucor circinelloidesand Penicillium aurantiogriseum\ mechanism of hy-drolysis. Arch Microbiol 169, 106-1 12.

35. Brimer, L., Cicalini, A. R., Federici, F. & Petruccioli,M. (1998). Microbial linamarases for the detoxifica-tion of vegetable products. Trop Agric 75, 166-168.

36. Brimer, L., Nout, M. J. R. & Tuncel, G. (1998). p-gly-cosidase (amygdalase and linamarase) from Endomycesfibuliger (LU677): formation and crude enzyme proper-ties. Appl Microbiol Biotechnol 49, 1 82-1 88.

37. Brimer, L. & Rosling, H. (1993). Microdiffusionmethod with solid state detection for determination ofcyanogenic glycosides from cassava in human urine.Food Chem Toxicolll, 599-603.

38. Brimer, L., Tuncel, G. & Nout, M. J. R. (1993). Simplescreening procedure for microorganisms to degradeamygdalin. Biotechnol Tech 7, 683-687.

39. Bruneton, J. (1995). Pharmacognosy, Phytochemistry,Medicinal Plants. Andover, United Kingdom:Lavoisier/Intercept.

40. Burnouf-Radosevich, M. & Delfel, N. E. (1984). High-performance liquid chromatography of oleane-typetriterpenes. JChromatogr 292, 403-409.

41. Burnouf-Radosevich, M. & Paupardin, C. (1983).Elaboration de saponines triterpeniques par de tissus deChenopodium quinoa Willd. cultives in vitro. C. R. Se-ances Acad Sd Ser III Sd Vie (Physiol Veget) 296,429-432.

42. Galloway, D. H., Hickey, C. A., & Murphy, E. L.(1971). Reduction of intestinal gas-forming propertiesof legumes by traditional and experimental food pro-cessing methods. J Food Sd 36, 251-255.

43. Carlson, L., Mlingi, N., Juma, A., Ronquist, G. &Rosling, H. (1999). Metabolic fates in humans oflinamarin in cassava flour ingested as stiff porridge.Food Chem Toxicol 37, 307-312.

44. Carter, J. H., McLafferty, M. A. & Goldman, P. (1980).Role of the gastrointestinal microflora in amygdalin(laetrile)-induced cyanide toxicity. BiochemPharmacol 29, 301-304.

45. Cereda, M. P. (1994). Processing of cassava roots inBrazil: safety implications. Acta Hortic 375, 225-226.

46. Cheeke, P. R. & Shull, L. R. (1985). Glycosides. InNatural Toxicants in Feeds and Poisonous Plants, pp.173-234. Westport, CT: AVI Publishing Company.

47. Chen, S. (1989). Preparation of fluid soymilk. In Pro-ceedings of the World Congress on Vegetable ProteinUtilization in Human Foods and Animal Feedstuffs,pp. 341-352. Edited by T. H. Applewhite. Champaign,IL: American Oil Chemists Society.

48. Chiwona-Karlrun, L., Mkumbira, J., Saka, J., Bovin, M.,Mahungu, N. M. & Rosling, H. (1998). The importanceof being bitter — a qualitative study on cassava cultivarpreference in Malawi. EcolFoodNutr 37, 219-245.

49. Cole, R. J., Cutler, H. G. & Stuart, B. P. (1989).Carboxyatractyloside. In Toxicants of Plant Origin.Vol. 2, Glycosides, pp. 253-263. Edited by P. R.Cheeke. Boca Raton, FL: CRC Press.

50. Conn, E. E. (1979). Cyanogenic glycosides. In Interna-tional Review of Biochemistry, Biochemistry of Nutri-tion IA, Vol. 27, pp. 21^3. Edited by A. Neuberger &T. H. Jukes. Baltimore, MD: University Park Press.

5 1 . Conn, E. E. ( 1 98 1 ). Unwanted biological substances infoods: cyanogenic glycosides. In Impact of Toxicologyon Food Processing (IFT Basic Symposium Series).Westport, CT: AVI Publishing Company.

52. Conn, E. E. (1991). The metabolism of a natural prod-uct: lessons learned from cyanogenic glycosides.Planta Med 57 (Suppl. 1), S1-S9.

53. Cooper, M. R. & Johnson, A. W. (1998). PoisonousPlants and Fungi in Britain (Animal and Human Poi-soning), 2nd edn. London: The Stationary Office.

54. Corchia, C., Balata, A., Meloni, G. F. & Meloni, T.(1995). Favism in a female newborn infant whosemother ingested fava beans before delivery. J Pediatr127, 807-808.

55. Crivelli, G., Torreggiani, D., Senesi, E., Forni, E.,Bertolo, G. & Maestrelli, A. (1991). Researches on theosmotic dehydration of apricots. Acta Hortic 293, 657-666.

56. Cruz, R., Batistela, J. C. & Wosiacki, G. (1982). Mi-crobial a-galactosidase for soymilk processing. J FoodSd 46, 1196-1200.

57. Dahlgren, R. M. T. (1980). A revised system of classi-fication of the angiosperms. BotJLinn Soc 80, 91—124.

58. Dahlgren, R., Rosendal-Jensen, S. & Nielsen, B. J.(1981). A revised classification of the angiospermswith comments on correlation between chemical andother characters. In Phytochemistry and AngiospermPhytogeny, pp. 149-204. Edited by D. A. Young & S.Seigler. New York: Praeger Publishers.

59. Dalsgaard, K., Hilgers, L. & Trouve, G. (1990). Classi-cal and new approaches to adjuvant use in domesticfood animals. Adv Vet Sd Comp Med 35, 121-160.

60. Delange, F., Ekpechi, L. O. & Rosling, H. (1994). Cas-sava cyanogenesis and iodine deficiency disorders.Acta Hortic 375, 289-293.

61 . Dendy, D. A. V. (1995). Sorghum and the millets: pro-duction and importance. In Sorghum and Millets,Chemistry and Technology, pp. 1 1—26. Edited by D. A.V. Dendy. St. Paul, MN: American Association of Ce-real Chemists.

62. Deshpande, S. S. (1992). Food legumes in human nu-trition: a personal perspective. CRC Crit Rev Food SdNutr 32, 333-363.

63. Dibofori, A. N., Okoh, P. N. & Onigbinde, A. O.(1994). Effect of germination on the cyanide and oli-gosaccharide content of lima beans (Phaseoluslunatus). Food Chem 51, 133-136.

64. Dirar, H. A. (1993). The Indigenous Fermented Foodsof the Sudan, a Study in African Food and Nutrition.Wallingford, United Kingdom: CAB International.

65. Doggett, H. (1988). Sorghum. London: Longman.

66. Due, G., Sixdenier, G., LiIa, M. & Furstoss, V. (1989).Search of genetic variability for vicine and convicinecontent in Vidafaba L.: a first report of a gene whichcodes for nearly zero- vicine and zero-convicine con-tents. In Recent Advances of Research inAntinutritional Factors in Legume Seeds, pp. 305-3 13.Edited by J. Huisman, T. F. B. van der Poel & I. E.Liener. Wageningen, Netherlands: Pudoc.

67. Dufour, D. L. (1988). Cyanide content of cassava(Manihot esculenta, Euphorbiaceae) cultivars used byTukanoan Indians in Northwest Amazonia. Econ Bot42, 255-266.

68. Dufour, D. (1989). Effectiveness of cassava detoxifica-tion techniques used by indigenous peoples in northwest Amazonia. Interciencia 14, 86-91.

69. Dufour, D. L. (1993). The bitter is sweet: a case studyof bitter cassava (Manihot esculenta) in Amazonia. InTropical Forests, People and Food, Biocultural Inter-actions and Applications to Development, pp. 575-588. Edited by C. M. Hladik, A. Hladik, O. F. Linares,H. Pagezy, A. Semple & M. Hadley. Paris/Carnforth:Unesco/The Parthenon Publishing Group.

70. Dunstan, W. R., Henry, T. A. & AuId, S. J. M. (1906).Cyanogenesis in plants V. the occurrence ofphaseolutin in cassava (Manihot aipi and Manihotutilissimd). Proc R Soc Lond SER B Biol SCI 78, 1 52-158.

71. Dziuk, H. E., Duke, G. E., Buck, R. J. & Janni, K. A.(1985). Digestive parameters in young turkeys fedyucca saponin. Poult Sd 64, 1 143-1 147.

72. Eggum, B. O., Jensen, S. K. & Soerensen, H. (1993).The nutritive value of dehulled protein rich rapeseedmeal produced by aqueous enzymatic extraction. BullGCIRC 9, 37^1.

73. Ekanayake, I. J. & Bokanga, M. (1995). A Review onProduction, Agronomy and Cyanogenesis. The Cas-sava Biotechnology Network: Proceedings of the Sec-ond International Scientific Meeting. Bogor, Indone-sia, August 22-26, 1994. Working Document no. 150.CaIi, Columbia: Centre International de AgriculturaTropical.

74. Errol, R. J., Nikawela, J. K., Gooneratne, J. &Theivendirarajah, K. (1994). Studies on the bitter prin-ciple and debittering of Palmyrah fruit pulp. JSd FoodAgric65, 185-189.

75. Esaki, S., Ohishi, A., Katsumata, A., Sugiyama, N. &Kamiya, S. (1993). Synthesis of 8-mannopyranosyl-containing disaccharides and phenols as substrates forthe oc-L-mannosidase activity of commercialnaringinase. Biosci Biotechnol Biochem 57, 2009-2103.

76. Espinoza, O. B., Perez, M. & Ramirez, M. S. (1992).Bitter cassava poisoning in eight children: a case re-port. Vet Hum Toxicol 34, 65.

77. Essers, A. J. A. (1995). Removal of cyanogens fromcassava roots; studies on domestic sun-drying andsolid-substrate fermentation in rural Africa. PhD diss.,Agricultural University Wageningen, Wageningen, theNetherlands.

78. Essers, A. J. A., Alsen, P. & Rosling, H. (1992). Insuf-ficient processing of cassava induced acute intoxica-tions and the paralytic disease konzo in a rural area ofMozambique. Ecol Food Nutr 27, 17-27.

79. Essers, A. J. A., Bennik, M. H. J. & Nout, M. J. R.(1995). Mechanism of increased linamarin degradationduring solid-substrate fermentation of cassava. WorldJ Microbiol Biotechnol 11, 266-270.

80. Essers, A. J. A., Ebong, C., Van der Grift, R. M., Nout,M. J. R., Otim-Nape, W. & Rosling, H. (1995). Reduc-ing cassava toxicity by heap-fermentation in Uganda.Int J Food Sd Nutr 46, 125-136.

81. Essers, A. J. A., Jurgens, C. M. G. A. & Nout, M. J. R.(1995). Contribution of selected fungi to the reductionof cyanogen levels during solid-substrate fermentationof cassava. Int J Food Microbiol 26, 251-257.

82. Evans, W. C., Patel, M. C. & Koohy, Y. (1982). Acutebracken poisoning in homogastric and ruminant ani-mals. Proc R. Soc Edinb Sect B 8 1 , 29-64.

83. Eyjolfsson, R. (1970). Recent advances in the chemis-try of cyanogenic glycosides. Fortschr Chem OrgNaturst 27, 74-108.

84. Fanourakis, N. E. & Tzifaki, E. E. (1993). Correlatedinheritance of fruit neck with fruit length and linkagerelations with 10 other characteristics of cucumber.Euphytica 65, 71-77.

85. Finnegan, T. J. A. & Lewis, M. J. (1985). Nitrogen ex-traction from defatted rapeseed, with particular refer-ence to United Kingdom commercial rapeseed meal. JSd FoodAgric 36, 520-530.

86. Finnemore, H. & Gledhill, W. C. (1928). The presenceof cyanogenetic glucosides in certain species of Aca-cia. Aust J Pharm 9, 174-178.

87. Fomunyam, R. T., Adegbola, A. A. & Oke, O. L.(1985). The stability of cyanohydrins. Food Chem 17,221-225.

88. Food and Agricultural Organization. (1995). Sorghumand Millets in Human Nutrition (F AO Food and Nutri-tion Series, No. 27). Rome: Author.

89. Food and Agricultural Organization. (1996). YearBook of Food Production. Rome: Food and Agricul-tural Organization.

90. Frehner, M., Scalet, M. & Conn, E. E. (1990). Patternof the cyanide-potential in developing fruits. PlantPhysiol 94, 28-34.

91. Fulmer, R. W. (1989). The preparation and propertiesof defatted soy flours and their products. In Proceed-ings of the World Congress on Vegetable Protein Utili-zation in Human Foods and Animal F eedstuffs, pp. 55-

61. Edited by T. H. Applewhite. Champaign, IL:American Oil Chemists Society.

92. Gaffield, W. & Keeler, R. F. (1994). Plant steroidal al-kaloid teratogens: structure-activity relations and im-plications. In Plant-Associated Toxins, Agricultural,Phytochemical and Ecological Aspects, pp. 333-338.Edited by S. M. Colegate & P. R. Dorling.Wallingford, United Kingdom: CAB International.

93. Galiano, S., Gaetani, G. F., Barabino, A., Cottafava, F.,Zeitlin, H., Town, M. & Luzzatto, L. (1990). Favism inthe African type of glucose-6-phosphate dehydroge-nase deficiency (A-). BriMedJ30Q, 236.

94. Garcia, V. V. & Palmer, J. K. (1980). Carbohydrates ofwinged beans, Psophocarpus tetragonologus (L.). JFood Techno! 15, 417-484.

95. Gee, J. M., Price, K. R., Ridout, C. L., Wortley, G. M.,Hurrell, R. F. & Johnson, I. T. (1993). Saponins ofQuinoa (Chenopodium quinoa): effect of processingon their abundance in quinoa products and their bio-logical effects on the intestinal mucosal tissue. J SdFoodAgric 63, 201-209.

96. Gelder, van W. M. J. (1991). Chemistry, toxicology,and occurrence of steroidal glycoalkaloids: potentialcontaminants of the potato (Solanum tuberosum L.). InPoisonous Plant Contamination of Edible Plants, pp.117-156. Edited by Abdel-Fattah M. Rizk. BocaRaton, FL: CRC Press.

97. Giraud, E., Gosselin, L. & Raimbault, M. (1993). Pro-duction of a Lactobacillus plantarum starter withlinamarase and amylase activities for cassava fermen-tation. JSd FoodAgric 62, 77-82.

98. Gonzalez, J. A., Estevez-Braun, A., Estevez-Reyes, R.,Bazzocchi, I. L., Moujir, L., Jimenez, I. A., Ravelo, A.G. & Gonzalez, A. G. (1995). Biological activity ofsecondary metabolites from Bupleurum salidfolium(Umbelliferae). Experientia 51, 35—39.

99. Grasselly, C. & Crossa-Raynaud, P. (1980).L'amandier, Techniques Agricoles et ProductionsMediterraneennes. Paris: G. -P. Maisonneuve etLarose.

100. Haberer, J. P., Jouve, P., Bedock, B. & Bazin, P. E.(1984). Severe hypokalaemia secondary to overindul-gence in alcohol-free pastis. Lancet 8376 (March 25),575-576.

101. Halkier, B. A. & Moeller, B. L. (1990). The biosynthe-sis of cyanogenic glucosides in higher plants. J BiolChem 265, 21114-21121.

102. Hampl, J. S., Holland, K. A., Marple, J. T., Hutchins,M. R. & Brockman, K. K. (1997). Acute hemolysis re-lated to consumption of fava beans: a case study andmedical nutition therapy approach. J Am Diet Assoc97, 182-183.

103. Harborne, J. B. & Baxter, H., eds. (1995). Phytochemi-cal Dictionary, a Handbook ofBioactive Compoundsfrom Plants. London: Taylor and Francis.

104. Hartemink, R., ed. (1998). Non-Digestible Oligosac-charides: Healthy Food for the Colon? Proceedings ofthe International Symposium, Wageningen, the Neth-erlands, 4-5 Dec., 1997. Wageningen, Netherlands:Wageningen Pers.

105. Hasegawa, S. & Maier, V. P. (1990). Biochemistry oflimonoid citrus juice bitter principles and biochemicaldebittering processes. In Bitterness in Foods and Bev-erages (Developments in Food Science 25), pp. 293-308. Edited by R. L. Rouseff. Amsterdam: Elsevier.

106. Hegazy, M. I. & Marquardt, R. R. (1983). Develop-ment of a simple procedure for the complete extractionof vicine and convicine from faba beans (Vicia fabaL.). J Sd Food Agric 34, 100-108.

107. Reiser, C. B. & Nelson, D. C. (1974). On the origin ofthe cultivated chenopods (Chenopodium). Genetics 78,503-505.

108. Herbert, V. (1979). Laetrile: the cult of cyanide: pro-moting poison for profit. Am J CHn Nutr 32, 1121-1158.

109. Hernandez, J. L., Adris, J., de Rank, E. F., Farias, R. &Samman, N. (1981). Predigested soybeans. J Am OilCfewiSbc 58, 510-511.

1 10. Hernandez, T., Lundquist, P., Oliveira, L., Crista, R. P.,Rodriguez, E. & Rosling, H. (1995). Fate in humans ofdietary intake of cyanogenic glycosides from roots ofsweet cassava consumed in Cuba. Nat Tox 3, 1 14-1 17.

111. Hirono, I. (1989). Carcinogenic bracken glycosides. InToxicants of Plant Origin. Vol. 2, Glycosides, pp. 239-251. Edited by P. R. Cheeke. Boca Raton, FL: CRCPress.

112. Hirono, L, Kono, Y., Takahashi, K., Yamada, K.,Niwa, H., Ojika, M., Kigoshi, H., Niiyama, K. &Uosaki, Y. (1984). Reproduction of acute bracken poi-soning in a calf with ptaquiloside, a bracken constitu-ent. Vet Rec 11 5, 375-378.

113. Hooper, P. T. (1978). Cycad poisoning in Australia —etiology and pathology. In Effects of Poisonous Plantson Livestock, pp. 337-347. Edited by R. F. Keeler, K.R. Van Kampen & L. F. James. New York: AcademicPress.

114. Hosel, W. (1981). Glycosylation and glycosidases. InThe Biochemistry of Plants, a Comprehensive Treatise.Vol. 7, Secondary Plant Products, pp. 725-753. Editedby P. K. Stumpf & E. E. Conn. London: AcademicPress.

115. Hewlett, W. P. (1994). Konzo: a new human diseaseentity. Acta Hortic 375, 323-329.

1 16. Hughes, A. (1981). The genetic control of plant cyano-genesis. In Cyanide in Biology, pp. 495-508. Edited byB. Vennesland, E. E. Conn, C. J. Knowles, J. Westley& F. Wissing. New York: Academic Press.

1 17. Hughes, A. & Conn, E. E. (1976). Cyanoglucoside bio-synthesis in white clover (Trifolium repens}. Phy-tochemistry 15, 697-701.

118. Hughes, A. & Maher, E. P. (1973). Studies on the na-ture of the Li locus in Trifolium repens L. I. purifica-tion and properties of the enzyme component. BiochemGenet 8, 1-12.

1 19. Hughes, M. A., Sharif, A. L., Dunn, M. A. & Oxtoby,E. (1988). The molecular biology of cyanogenesis. InCyanide Compounds in Biology; CIBA FoundationSymposium No. 140, pp. 111-130. Edited by DavidEvered and Sarah Harnett. Chichester, England: JohnWiley & Sons.

120. Hulse, J. H., Laing, E. M. & Pearson, O. E. (1980).Nutritional inhibitors and toxic factors. In Sorghumand the Millets: Their Composition and NutritiveValue, pp. 297-395. London: Academic Press.

121. Ikediobi, C. O. & Onyike, E. (1982). The use oflinamarase in gari production. Process Biochem 17, 2-5.

122. Ingvorsen, K., Yde, B., Godtfredsen, S. E. & Tsuchiya,R. T. (1988). Microbial hydrolysis of organic nitrilesand amides. In Cyanide Compounds in Biology; CIBAFoundation Symposium No. 140, pp. 16-31. Edited byDavid Evered and Sarah Harnett. Chichester, England:John Wiley & Sons.

123. Jadhav, S. J., Sharma, R. P. & Salunke, D. K. (1981).Naturally occurring alkaloids in foods. CRC Crit RevToxicol 9, 2 1-1 04.

124. Jansz, E. R. & Pieris, N. (1978). Studies on some locallegumes. II. Cyanogenic glucosides. J Nat Sd CouncSri Lanka 6, 1-9.

125. Joint FAO/WHO Expert Committee on Food Addi-tives. (1986). Evaluation of certain food additives andcontaminants: twenty-ninth report of the joint FAO/WHO Expert Committee on Food Additives, WHOTech Rep Ser 733 ,41.

126. Johns, T. (1990). With Bitter Herbs They Shall Eat It:Chemical Ecology and the Origins of Human Diet andMedicine. Tucson, AZ: The University of ArizonaPress.

127. Johns, T. (1994). Defence of nitrogen-rich seeds con-strains selection for reduced toxicity during the domes-tication of the grain legumes. In Advances in LegumeSystematics. Part 5, The nitrogen Factor, pp. 151—167.Edited by J. I. Sprent & D. M. McKey. London: TheRoyal Botanic Gardens.

128. Johnson, I. T., Gee, J. M., Price, K. R. & Fenwick, G.R. (1989). Gastrointestinal effects of somemembranolytic plant constituents. In Recent Advancesof Research in Antinutritional Factors in LegumeSeeds, pp. 206-209. Edited by J. Huisman, T. F. B. vander Poel, & I. E. Liener. Wageningen, Netherlands:Pudoc.

129. Jones, D. A. (1998). Why are so many food plants cya-nogenic? Phytochemistry 47, 155-162.

130. Kaplan, M. A. C., Figueiredo, M. R. & Gottlieb, O.(1983). Variation in cyanogenesis in plants with seasonand insect pressure. Biochem Syst Ecol 11, 367-370.

131. Kellerman, T. S., Miles, C. O., Erasmus, G. L.,Wilkins, A. L. & Coetzer, J. A. W. (1994). The pos-sible role of steroidal saponins in the pathogenesis ofGeeldikkop, a major hepatogenous photosensitation ofsmall stock in South Africa. In Plant-Associated Tox-ins, Agricultural, Phytochemical and Ecological As-pects, pp. 287-292. Edited by S. M. Colegate & P. R.Dorling. Wallingford, United Kingdom: CAB Interna-tional.

132. Kensil, C. R. (1996). Saponins as vaccine adjuvants.Cr Rev Therap Drug Carr Syst 13, 1-55.

133. Kerlan, V. Ogor, C. & Bercovici, J. P. (1994). Intoxica-tion a Ia glycyrrhizine apres un sevrage tabagique.PresseMed23(l\5Q.

134. Ketiku, A. O., Akinyele, I. O., Keshiro, O. O. &Akinnawo, O. O. (1978). Changes in the hydrocyanicacid concentration during traditional processing of cas-sava into "gari" and "lafun." Food Chem 3, 221-228.

135. King, N. L. R. & Bradbury, J. H. (1995). Bitterness ofcassava: identification of a new apiosyl glucoside andother compounds that affect its bitter taste. J Sd FoodAgric 68, 223-230.

136. Klopfenstein, C. F. & Hoseney, R. C. (1995). Nutri-tional properties of sorghum and the millets. In Sor-ghum and Millets, Chemistry and Technology, pp.125-168. Edited by D. A. V. Dendy. St. Paul, MN:American Association of Cereal Chemists.

137. Kloss, P., Jeffery, E., Tumbleson, M., Zhang, Y., Par-sons, C. & Wallig, M. (1996). Studies on the toxic ef-fects of crambe meal and two of its constituents, 1-cy-ano-2-hydroxy-3-butene (CHB) and epi-progoitrin, inbroiler chick diets. Br Poult Sci 37, 971-986.

138. Knowles, C. J. (1988). Cyanide utilization and degra-dation by microorganisms In Cyanide Compounds inBiology; CIBA Foundation Symposium No. 140, pp. 3-15. Edited by David Evered and Sarah Harnett.Chichester, England: John Wiley & Sons.

139. Koziol, M. J. (1991). Afrosimetric estimation ofthreshold saponin concentration for bitterness inquinoa (Chenopodium quinoa Willd). J Sd Food Agric54,211-219.

140. Kuiper-Goodman, T. & Nawrot, P. S. (1993). Solanineand chaconine. WHO FoodAddit Ser 30, 339-372.

141. Kupchella, L. & Syty, A. (1984). Determination ofcyanogenic glycosides in seeds by molecular absorp-tion spectrometry in the gas phase. J Assoc Off AnalChem 67, 188-191.

142. Lancaster, P. A. & Brooks, J. E. (1983). Cassava leavesas human food. Econ Bot 37, 331-348.

143. Lancaster, P. A., Ingham, J. S., Lim, M. Y. & Coursey,D. G. (1982). Traditional cassava-based foods: surveyof processing techniques. Econ Bot 36, 12-45.

144. Lansky, P. S. (1993). Plants that lower cholesterol.Acta Hortic 332, 131-136,

145. Lea, A. G. H. (1991). Enzymes in the production ofbeverages and fruit juices. In Enzymes in Food Pro-cessing, pp. 194-220. Edited by G. A. Tucker & L. F. J.Woods. Glasgow, Scotland: Blackie Academic andProfessional.

146. Lechtenberg, M. & Nahrstedt, A. (1999). Cyanogenicglycosides. In Naturally Occurring Glycosides, pp.147-191. Edited by R. Ikan. Chichester, England: JohnWiley & Sons.

147. Legras, J. L., Jory, M., Arnaud, A. & Galzy, P. (1990).Detoxification of cassava pulp using Brevibacteriumsp. R3U.Appl Microbiol Biotechnoni, 529-533.

148. Legras, J. L., Kaakeh, M. R., Arnaud, A. & Galzy, P.(1989). Degradation of cyanoglucosides byBrevibacterium sp. R312 strain. J Gen Appl Microbiol35,451^*61.

149. Lei, V., Amoa-Awua, W. K. A. & Brimer, L. (1999).Degradation of cyanogenic glycosides by Lactobacil-lus plantarum strains from spontaneous cassava fer-mentation and other microorganisms. Int J FoodMicrobiol 53, 169-184.

150. Lewin, R. (1987). Environmental hypothesis for braindiseases strengthened by new data. Science 237, 483-484.

151. Liener, I. E. (1987). Detoxifying enzymes. In FoodBiotechnology. Vol. 1, pp. 249-271. Edited by R. D.King & P. S. J. Cheetham. London: Elsevier AppliedScience.

152. Loew, D., von Bergmann, U., Schmidt, M. & Uberla,K. H. (1994). Anthranoidlaxanzien. Ursache furkolonkarzinom? Dtsch Apoth Ztg 1 34, 3 1 80-3 1 83 .

153. Low, V. H. K. & Thomson, J. A. (1990). Cyanogenesisin Australian bracken (Pteridium esculentum): distri-bution of cyanogenic phenotypes and factors influenc-ing activity of the cyanogenic glucosidase. In BrackenBiology and Management, pp. 105-111. Edited by J.A. Thomson & R. T. Smith. Wahroonga, NSW Austra-lia: The Australian Institute of Agricultural Science.

154. Luca, P., de Moretti, A., Sabato, S. & Gigliano, S. S.(1980). The ubiquity of cycasin in cycads. Phytochem-istry 19, 2230-2231.

155. Luckner, M. (1990). Secondary Metabolism in Micro-organisms, Plants, and Animals, 3rd edn. Verlag, Jena:VEB Gustav Fischer.

156. Madhusudhan, K. T. & Singh, N. (1985). Effect ofdetoxification on the physicochemical properties oflinseed. J Agric Food Chem 33, 12-19.

157. Madhusudhan, K. T. & Singh, N. (1985). Effect of heattreatment on the functional properties of linseed meal.J Agric Food Chem 33, 19-22.

158. Maharaj, L, Froh, K. J., & Campbell, J. B. (1986). Im-mune responses of mice to inactivated rabies vaccineadministered orally: potentiation by Quillaja saponin.Can J Microbiol 32, 414^120.

159. Maher, E. P. & Hughes, A. (1973). Studies on the na-ture of the Li locus in Trifolium repens L. II. the effectof genotype on enzyme activity and properties.Biochem Genet 8, 13-26.

160. Majak, W. & Pass, M. A. (1989). Aliphaticnitrocompounds. In Toxicants of Plant Origin. Vol. 2,Glycosides, pp. 143-159. Edited by P. R. Cheeke.Boca Raton, FL: CRC Press.

161. Malcolm, S. B. (1991). Cardenolide-mediated interac-tions between plants and herbivores. In Herbivores:Their Interactions with Secondary Plant Metabolites,2nd edn. Vol. 1, The Chemical Participants, pp. 251-296. Edited by G. A. Rosenthal & M. R. Berenbaum.San Diego: Academic Press.

162. Marquardt, R. R. & Arbid, M. S. S. (1988). Protectionagainst the toxic effects of the favism factor (divicine)in rats by vitamins E, A and C and iron chelatingagents. JSd FoodAgric 43, 155-166.

163. Matsushima, T., Matsumoto, H., Shirai, A., Sawamura,M. & Sugimura, T. (1979). Mutagenicity of the natu-rally occurring carcinogen cycasin and syntheticmethylazoxymethanol conjugates in Salmonellatyphimurium. Cancer Res 39, 3780-3782.

164. McKay, A. M. (1992). Hydrolysis of vicine andconvicine from faba beans by microbial p-glucosidaseenzymes. J Appl Bacteriol 72, 475^78.

165. McMahon, J. M., White, W. L. B. & Sayre, R. T.(1995). Cyanogenesis in cassava (Manihot esculentaCrantz). JExp Bot 46, 73 1-741 .

166. McMillan, D. C., Schey, K. L., Meier, G. P. & Jollow,D. J. (1993). Chemical analysis and hemolytic activityof the fava bean aglycone divicine. Chem Res Toxicol6, 439-444.

167. Meijer, M. M. T. & Muuse, B. G. (1989).Optimalization of dehulling technique and enzymatichydrolysis of vicine/convicine to eliminate ANF 's ofFaba bean. In Recent Advances of Research inAntinutritional Factors in Legume Seeds, pp. 268-271 .Edited by J. Huisman, T. F. B. van der Poel, & I. E.Liener. Wageningen, Netherlands: Pudoc.

168. Mickelsen, O., Campbell, M. E., Yang, M., Mugera, G.M. & Whitehair, C. K. (1964). Studies with cycad. FedProc23, 1363-1365.

169. Ministry of Health Mozambique. (1984). Mantakassa:an epidemic of spastic paraparesis associated withchronic cyanide intoxication in a cassava stable area inMozambique. Bull WHO 62, 477-492.

170. Mizutani, F., Hirota, R., Amano, S., Hino, A. &Kadoya, K. (1991). Changes in cyanogenic glycosidecontent and (3-cyanoalanine synthetase activity in fleshand seeds of Japanese plum (Prunus salicina Lindl.)during development. J Jpn Soc Hortic Sd 59, 863-867.

171. Mkpong, O. E., Yan, H., Chism, G. & Sayre, R. T.(1990). Purification, characterization, and localizationof linamarase in cassava. Plant Physiol 93, 1 76-1 8 1 .

172. Mlingi, N., Poulter, N. H. & Rosling, H. (1992). Anoutbreak of acute intoxications from human consump-tion of insufficiently processed cassava in Tanzania.NutrRes 12, 677-687.

173. Moertel, C. G., Ames, M. M., Kovach, J. S., Mayer, T.P., Rubin, J. R. & Tinker, J. A. (1981). A pharmaco-logic and toxicological study of amygdalin. JAm MedAssoc 245, 48-50.

174. Monastra, F. & Fideghelli, C. (1980). L'albicocco.Agricoltura e Ricerca 10, 24—33.

175. Moncrieff, R. W. (1967). The Chemical Senses. Cleve-land, OH: CRC Press.

176. Monzini, A. & Gorini, F. (1991). Pre-freezing, storageand industrial use of apricot with relationship to qual-ity. Acta Hortic 293, 599-655.

177. Moretti, A., Sabato, S. & Gigliano, G. S. (1981). Dis-tribution of macrozamin in australasian cycads. Phy-tochemistry2Q, 1415-1416.

178. Morgan, R. W. & Hoffman, G. R. (1983). Cycasin andits mutagenic metabolites. Mutat Res 1 14, 19-58.

179. Munro, J. M. (1987). Cotton, 2nd edn. Harlow, UnitedKingdom: Longman Scientific and Technical.

180. Nachman, R. J. & Olson, J. D. (1983). Ranunculin. atoxic constituent of the poisonous range plant bur but-tercup (Ceratocephalus testiculatus). J Agric FoodChem 31, 1358-1360.

181. Nafa, K., Reghis, A., Osmani, N., Baghli, L., Ait- Ab-bes, H., Benabadji, M., Kaplan, J. -C., Vulliamy, T. &Luzzatto, L. (1994). At least five polymorphic mutantsaccount for the prevalence of glucose-6-phosphate de-hydrogenase deficiency in Algeria. Hum Genet 94,513-517.

182. Nahrstedt, A. (1987). Recent developments in chemis-try, distribution and biology of the cyanogenic glyco-sides. In Biologically Active Natural Products, pp.213-234. Edited by K. Hostettmann & P. J. Lea. Ox-ford: Oxford University Press.

183. Nartey, F. (1968). Studies on cassava, Manihotutilissima Pohl. I. cyanogenesis: the biosynthesis oflinamarin and lotaustralin in etiolated seedlings. Phy-tochemistryl, 1307-1312.

184. Nartey, F. (1978). Manihot esculenta (Cassava), cy-anogenesis, ultrastructure and seed germination.Copenhagen: Munksgaard.

185. Nout, M. J. R., Tuncel, G. & Brimer, L. (1995). Micro-bial degradation of amygdalin of bitter apricot seeds(Punus armeniaca). Int J Food Microbiol 24, 407—412.

186. O'Brien, G. M., Wheatley, C. C., Iglesias, C. &Poulter, N. H. (1994). Evaluation, modification andcomparison of two rapid assays for cyanogens in cas-sava. JSd FoodAgric 65, 391-399.

1 87. Oelrichs, P. B., Pearce, C. M., Kudo, K. & Kelly, W. R.(1994). The isolation, structure elucidation and toxicityof the kaurene glycosides parquin and carboxyparquin

in Cestrum parqui. . In Plant-Associated Toxins, Agri-cultural, Phytochemical and Ecological Aspects, pp.251-255. Edited by S. M. Colegate & P. R. Dorling.Wallingford, United Kingdom: CAB International.

188. Ohga, S., Higashi, E., Nomura, A., Matsuzaki, A.,Hirono, A., Miwa, S. & Fujii, H. (1995). Haptoglobintherapy for acute favism: a Japanese boy with glucose-6-phosphate dehydrogenase Guadalajara. Br JHaematol 89, 421^123.

189. Ohtsubo, T. & Ikeda, F. (1994). Seasonal changes ofcyanogenic glycosides in Mume (Prunus mume Sieb.et Zucc.) seeds. J Jpn Soc Hortic Sd 62, 695-700.

190. Ojika, M., Sugimoto, K., Okazaki, T. & Yamada, K.(1989). Modification and cleavage of DNA byptaquiloside: a new potent carcinogen isolated frombracken fern. J Chem Soc Chem Commun 22, 1775-1777.

191. Okafor, N. & Ejiofor, A. O. (1990). Rapid detoxifica-tion of cassava mash fermenting for gari productionfollowing the inoculation with a yeast simultaneouslyproducing linamarase and amylase. Process Biochem25, 82-86.

192. Okolie, N. P. & Ugochukwu, E. N. (1989). Cyanidecontents of some Nigerian legumes and the effect ofsimple processing. Food Chem 32, 209-216.

193. Okuyama, T., Takata, M., Takahashi, K., Ishikawa, T.,Miyasaka, K. & Kaneyama, N. (1989). High-perfor-mance liquid chromatographic analysis of naturallyoccurring glycosides and saponins. J Chromatogr 466,390-398.

194. Olafsdottir, E. S., Andersen, J. V. & Jaroszewski, J. W.(1989). Cyanohydrin glycosides of Passifloraceae.Phytochemistry 28, 127-132.

195. Olson, J. D., Anderson, T. E., Murphy, J. C. & Madsen,G. (1983). Bur buttercup poisoning of sheep. JAm VetMed Assoc 183, 538-543.

196. O'Mullane, J. E. & Hayter, I. P. (1993). InternationalPatent PCT/GB93/00343.

197. Onabolu, A. O., Bokanga, M. & Rosling, H. (1999).Cassava processing in a Nigerian community affectedby a neuropathy attributed to dietary cyanide exposure.Trop Sd 39, 129-135.

198. Oomah, B. D. & Mazza, G. (1995). Functional proper-ties, uses of flaxseed protein. Inform 6, 1246-1252.

199. Oomah, B. D., Mazza, G. & Cui, W. (1994). Optimiza-tion of protein extraction from flaxseed meal. Food ResInt 27, 355-361.

200. Oomah, B. D., Kenaschuk, E. O., Cui, W. & Mazza, G.(1995). Variation in the composition of water-solublepolysaccharides in flaxseed. J Agric Food Chem 43,1484-1488.

201. Oomah, B. D., Mazza, G. & Kenaschuk, E. O. (1992).Cyanogenic compounds in flaxseed. J Agric FoodChem 40, 1346-1348.

202. Padmaja, G. (1995). Cyanide detoxification in cassavafor food and feed uses. CRC Cri Rev Food Sd Technol35, 299-339.

203. Palter, R., Lundin, R. E. & Haddon, W. F. (1972). Acathartic lignan glycoside isolated from CarthamusTinctorus L., Compositae. Phytochemistry 11, 2871-2874.

204. Pass, M. A. (1994). Toxicity of plant-derived aliphaticnotrotoxins. In Plant-Associated Toxins, Agricultural,Phytochemical and Ecological Aspects, pp. 541-545.Edited by S. M. Colegate & P. R. Dorling.Wallingford, United Kingdom: CAB International.

205. Pederson, G. A. & Brink, G. E. (1998). Cyanogenesiseffect on insect damage to seedling white clover in aBermuda grass sod. Agron /90, 208-210.

206. Petersen, H. W., Moelgaard, P., Nyman, U. & Olsen,C. E. (1993). Chemotaxonomy of the tuber-bearingSolanum species, subsection Potatoe (Solanaceae).Biochem Biophys Ecol2\, 629-644.

207. Petruccioli, M., Bimer, L., Cicalii, A. R. & Federici, F.(1999). The linamarase of Mucor drcinelloides LUM40 and its detoxifying activity on cassava. J ApplMicrobiolS6, 302-310.

208. Philbrick, D. J., Hill, D. C. & Alexander, J. C. (1977).Physiological and biochemical changes associatedwith linamarin administration to rats. Appl Pharmacol42,539-551.

209. Pieters, L. A. C. & Vlietinck, A. J. (1991). Naturallyoccurring carcinogens in plant foodstufs. In PoisonousPlant Contamination of Edible Plants, pp. 1-25. Ed-ited by Abdel-Fattah M. Rizk. Boca Raton, FL: CRCPress.

210. Poulton, J. E. (1983). Cyanogenic compounds in plantsand their toxic effects. In Handbook of Natural Toxins,pp. 1 18-157. Edited by R. F. Keeler & A. T. Tu. NewYork: Marcel Dekker.

211. Poulton, J. E. & Li, C. P. (1994). Tissue levelcompartmentation of (R)-amygdalin and amygdalinhydrolase prevents large-scale cyanogenesis in undam-aged Prunus seeds. Plant Physiol 104, 29-35.

212. Puri, A. (1990). Removal of bitter compounds from cit-rus products by adsorption techniques. In Bitterness inFoods and Beverages (Developments in Food Science25), pp. 325-336. Edited by R. L. Rouseff.Amsterdam: Elsevier.

213. Pusztai, A., Ewen, S. W. B., Grant, G., Brown, D. S.,Stewart, J. C., Peumans, W. J., van Damme, E. J. M. &Bardicz, S. (1993). Antinutritive effects of wheat-germagglutinin and other N-acetylglucosamine-specificlectins. BrJNutr 70, 313-321.

214. Rabot, S., Nugon-Baudon, L., Raibaud, P. & Szylit, O.(1993). Rape-seed meal toxicity in gnotobiotic rats: in-fluence of a whole human faecal flora or single humanstrains of Escherichia coli and Bacteriodes vulgatus.BrJNutr 70, 323-331.

215. Rachis, J. J. (1975). Oligosaccharides of food legumes:alpha-galactosidase activity and the flatus problem. InPhysiological Effects of Food Carbohydrates, pp. 207-222. Edited by A. Jeanes & J. Hodge. Washington, DC:American Chemical Society.

216. Raddick, J. G. (1989). The acetylcholinesterase-inhibi-tory activity of steroidal glycoalkaloids and their agly-cones. Phytochemistry 28, 2631-2634.

217. Raffauf, R. F. (1996). Plant Alkaloids: A Guide toTheir Discovery and Distribution. New York: FoodProducts Press.

218. Rahman, S. & Subrahmanyan, V. (1947). Effect of dif-ferent treatments on the removal of hydrocyanic acidfrom the Burma bean (P. lunatus Linn.). Curr Sd 11,351-352.

219. Ramsay, G. & Griffiths, D. W. (1996). Accumulationof vicine and convicine in Vicia faba and V.narbonensis. Phytochemistry 42, 63-67.

220. Rauws, A. G., Oiling, M. & Timmerman, A. (1982).The pharmacokinetics of amygdalin. Arch Toxicol 49,311-319.

221. Rauws, A. G., Oiling, M. & Timmerman, A. (1983).The pharmakokinetics of prunasin, a metabolite ofamygdalin. J Toxicol CHn Toxicol 19, 851-856.

222. Reddy, N. R. & Pierson, M. D. (1994). Reduction inantinutritional and toxic components in plant foods byfermentation. Food Res Int 27, 281-290.

223. Reichert, R. D., Tatarynovich, J. E. & Tyler, R. T.(1986). Abrasive dehulling of quinoa (Chenopodiumquinoa): effect on saponin content as determined by anadapted haemolytic assay. Cereal Chem 63, 471-475.

224. Reynolds, J. H. (1974). An immobilized oc-galactosi-dase continous flow reactor. Biotechnol Bioeng 16,135-147.

225. Rezakhani, A. & Maham, M. (1994). Cardiac manifes-tations of oleander poisoning in cattle and donkeys. InPlant-Associated Toxins, Agricultural, Phytochemicaland Ecological Aspects, pp. 534-537. Edited by S. M.Colegate & P. R. Dorling. Wallingford, United King-dom: CAB International.

226. Ridout, C. L., Price, K. R., DuPont, M. S., Parker, M.L. & Fenwick, G. R. ( 1991). Quinoa saponins — analy-sis and preliminary investigations into the effect of re-duction by processing. J Sd Agric Food 54, 165—176.

227. Robins, R. J., Rhodes, M. J. C., Parr, A. J. & Walton,N. J. (1990). Biosynthesis of bitter compounds. In Bit-terness in Foods and Beverages (Developments inFood Science 25), pp. 49-79. Edited by R. L. Rouseff.Amsterdam: Elsevier.

228. Rombouts, F. M. & Nout, M. J. R. (1995). Microbialfermentation in the production of plant foods. J ApplBacterioll9 (Symp. Suppl.), 108S-117S.

229. Romero, C., Manjon, A., Bastida, J. & Iborra, J. L.(1985). A method for assaying the rhamnosidase activ-ity of naringinase. Anal Biochem 149, 566-571.

230. Rosling, H. (1993). Cyanide exposure from linseed.Lancet 34 1,177.

23 1 . Rosling, H. (1994). Measuring effects in humans of di-etary cyanide exposure from cassava. Ada Hortic 375,271-283.

232. Rosling, H. & Tylleskar, T. (1996). Konzo. In TropicalNeurology, pp. 353-364. Edited by R. A. Shakin, P. K.Newman & C. M. Poser. London: W. B. SaundersCompany.

233. Ruiz-Teran, F. & Owens, J. D. (1999). Fate of oli-gosaccharides during production of soya bean tempe. JSd Food Agric 79, 249-252.

234. Saka, J. D. K., Mhone, A. R. K., Mkambira, J., Brimer,L., Bokanga, M., Mahungu, N. M., Chiwona-Karltun,L. & Rosling, H. (1998). Correlation between cyano-genic glucoside content and taste of fresh cassavaroots. Trop Agric 75, 169-173.

235. Sakata, M., Yoshida, A., Yuasa, C., Sakata, K. & Haga,M. (1987). Toxicity of D,L-mandelonitrile-p-D-gluco-side "prulaurasin" in rat. J Toxicol Sd 12, 47-55.

236. Salkowski, A. A. & Penney, D. G. (1994). Cyanidepoisoning in animals and humans: a review. Vet HumToxicol 36, 455^66.

237. Sanni, M. O., Sobaminwa, A. O., Eyinla, C. M. &Rosling, H. (1994). Safety aspects of processing cas-sava to gari in Nigeria. Ada Hortic 375, 227-23 1 .

238. Schab, R. & Yannai, S. (1973). An improved methodfor debittering apricot kernels. J Food Sd Technol 10,57-59.

239. Schambelan, M. (1994). Licorice ingestion and bloodpressure regulating hormones. Steroids 59, 127-130.

240. Schilcher, H. & Wilkens-Sauter, M. (1986). Quantita-tive determination of cyanogenic glycosides in Linumusitatissimum using HPLC. Fette, Seifen, Anstrichm 8,287-290.

241. Seigler, D. S. (1991). Cyanide and cyanogenic glyco-sides. In Herbivores: Their Interactions with Second-ary Plant Metabolites, 2nd edn, Vol. 1., pp. 35-77.Edited by G. A. Rosenthal & M. R. Berenbaum. SanDiego: Academic Press.

242. Seigler, D. S., Maslin, B. R. & Conn, E. E. (1989). Cy-anogenesis in the Leguminosae. In Advances in LegumeBiology, pp. 645-672. Edited by C. H. Stirton & J. L.Zarucchi. St. Louis, MO: Missouri Botanical Garden.

243. Shahidi, F. & Naczk, M. (1990). Removal ofglucosinolates and other antinutrients from canola andrapeseed by methanol/ammonia processing. In Canolaand Rapeseed; Production, Chemistry, Nutrition andProcessing Technology, pp. 291-306. Edited by F.Shahidi. New York: Van Nostrand Reinhold.

244. Shaw, P. E. (1990). Cyclodextrin polymers in the re-moval of bitter compounds in citrus juices. In Bitter-ness in Foods and Beverages (Developments in FoodScience 25), pp. 309-324. Edited by R. L. Rouseff.Amsterdam: Elsevier.

245. Sherba, S. E. (1970). South African Patent 6,902,503.

246. Shimizu, T., Yasuda, N., Kono, L, Yagi, F., Tadera, K.& Kobayashi, A. (1986). Hepatic and spinal lesions ingoats chronically intoxicated with cycasin. Jpn J VetSW 48, 1291-1295.

247. Sinha, S. K. & Nair, T. V. R. (1968). Studies on thevariability of cyanogenic glucoside content in cassavatubers. IndJAgric Sd 38, 958-963.

248. Smiley, K. L., Hensley, D. E. & Gasdorf, H. J. (1976).Alpha-galactosidase production and use in a hollow-fiber reactor. Appl Environ Microbiol 31,615-617.

249. Smith, B. L., Seawright, A. A., Ng, J. C., Hertle, A. T.,Thomson, J. A., & Bostock, P. D. (1994). Concentra-tion of ptaquiloside, a major carcinogen in bracken fern(Pteridium spp.), from eastern Australia and from acultivated worldwide collection held in Sydney, Aus-tralia. Nat Toxins 2, 347-353.

250. Smith, B. L., Shaw, G., Prakash, A. & Seawright, A. A.(1994). Studies on DNA adduct formation byptaquiloside, the carcinogen of bracken ferns(Pteridium spp.). In Plant-Associated Toxins, Agricul-tural, Phytochemical and Ecological Aspects, pp. 167-172. Edited by S. M. Colegate & P. R. Dorling.Wallingford, United Kingdom: CAB International.

251. Smith, C. R., Weisleder, D., Miller, R., Palmer, I. S. &Olson, O. E. (1980). Linustatin and Neolinustatin: cya-nogenic glycosides of linseed meal that protect animalsagainst selenium toxicity. J Org Chem 45, 507-510.

252. Soerensen, H. (1990). Glucosinolates: structure-prop-erties-function. In Canola and Rapeseed: Production,Chemistry, Nutrition and Processing Technology, pp.149-172. Edited by F. Shahidi. New York: VanNostrand Reinhold.

253. Speijers, G. (1993). Cyanogenic glycosides. WHOFood AdditSer 30, 299-337.

254. Spencer, P. S., Nunn, P. B., Hugon, J., Ludolph, A. C.,Ross, S. M., Roy, D. N. & Robertson, R. C. (1987).Guam amyotrophic lateral sclerosis-parkinsonism-de-mentia linked to a plant exitant neurotoxin. Science237,517-522.

255. Staron, T. (1974). La detoxification des tourteaux decolza par voie biologique. U alimentation et Ia Vie, 62,165-179.

256. Stoermer, F. C., Reistadt, R. & Alexander, J. (1993).Glycyrrhizic acid in liquorice: evaluation of healthhazard. Food Chem Toxicol31, 303-313.

257. Stoewsand, G. S. (1995). Bioactive organosulfurphytochemicals in Brassica oleracea vegetables: a re-view. Food Chem Toxic 33, 537-543.

258. Stone, R. (1993). Guam: deadly disease dying out. Sd-ence 26 1,424-426.

259. Strada, G. D., Fideghelli, C., Monastrata, F., Pennone,F. & Quarta, R. (1991). Apricot breeding at theInstitute Sperimentale per Ia Frutticoltura (ISF) ofRome. Acta Hortic 293, 121-132.

260. Sturm, W. (1982). Mandeln aus dem Mittelmeerraumund/oder Kalifornien. Vorkommen und geschmack-lische auswirkungen bitterer mandeln und diebedeutung rein sii(er mandeln. Susswaren 26, 196-198.

261. Sugimoto, H. & Van Buren, J. P. (1970). Removal ofoligosaccharides from soy milk by an enzyme fromAspergillus saitoi. J Food Sd 35, 655—660.

262. Sunnerheim-Sjoberg, K. (1991). Chemical studies ofsecondary metabolites in betula and pinus — with em-phasis on defence against mammalian herbivores. PhDdiss., Swedish University of Agricultural Sciences,Uppsala, Sweden.

263. Swai, A. B. M., Mclarty, D. G., Mtinangi, B. L., Tatala,S., Kitange, H. M., Mlingi, N., Rosling, H., Hewlett,W. P., Brubaker, G. R. & Albert!, K. G. M. M. (1992).Diabetes is not caused by cassava toxicity. DiabetesCare 15, 1378-1385.

264. Thananukul, D., Tanaka, N., Chichester, C. O. & Lee,T. -C. (1976). Degradation of raffmose and stachyosein soybean milk by oc-galactosidase from Mortierellevinecea: entrapment of galactosidase within polyacry-lamide gel. J Food Sd 41, 173-175.

265. Tjon Sie Fat, L. A. (1979). Contribution to the knowl-edge of cyanogenesis in angiosperms (CyanogeneVerbindingen bij Poaceae, Commelinaceae,Ranunculaceae en Campanulaceae). PhD diss., Uni-versity of Leiden, Leiden.

266. Tuncel, G., Nout, M. J. R. & Brimer, L. (1995). Theeffect of grinding, soaking and cooking on the degrada-tion of amygdalin of bitter apricot seeds. Food Chem53,447-451.

267. Tuncel, G., Nout, M. J. R. & Brimer, L. (1998). Degra-dation of cyanogenic glycosides of bitter apricot seeds(Prunus armeniacd) by endogenous and added en-zymes as affected by heat treatments and particle size.Food Chem 63, 65-69.

268. Tuncel, G., Nout, M. J. R., Brimer, L., & Goktan, D.(1990). Toxicological, nutritional and microbiologicalevaluation of tempe fermentation with Rhizopusoligosporus of bitter and sweet apricot seeds. Int JFood Microbiol, 11, 337-344.

269. Tylleskar, T. (1994). The association between cassavaand the paralytic disease konzo. Acta Hortic 375, 33 1-339.

270. Tylleskar, T., Banea, M., Bikangi, N., Cooke, R.,Poulter, N. & Rosling, H. (1992). Cassava cyanogensand konzo, an upper motoneuron disease found in Af-rica. Lancet 339, 208-21 1.

271. Vasconcelos, A. T., Twiddy, D. R., Westby, A. &Reilly, P. J. A. (1990). Detoxification of cassava dur-ing gari preparation. Int J Food Sd Tech 25, 198-203.

272. Verbiscar, A. J., Baniga, T. F., Weber, C. W., Reid, B.L., Swingle, R. S., Tei, J. E. & Nelson, E. A. (1981).Detoxification of jojoba meal by lactobacilli. J AgricFood Chem 29, 296-302.

273. Verjee, Z. H. (1993). Glucose-6-phosphate dehydroge-nase deficiency in Africa — a review. East Afr Med J70(4) supplement, 40^7.

274. Wallig, M. A., Kingston, S., Staack, R. & Jeffery, E. H.(1998). Induction of rat pancreatic glutathione S- trans-ferase and quinone reductase activities by a mixture ofglucosiolate breakdown derivatives found in brusselssprouts. Food Chem Toxicol 36, 365-373.

275. Wanasundara, P. K. J. P. D., Amarowicz, R., Kara, M.T. & Shahidi, F. (1993). Removal of cyanogenic glyco-sides of flaxseed meal. Food Chem 48, 263-266.

276. Wang, P. -X. & Ueberschar, K. H. (1990). The estima-tion of vicine, convicine and condensed tannins in 22varieties of faba beans (Viciafaba L.). Anim Feed SdTechnolll, 157-165.

277. Weber, C. W., Berry, J. W. & Cooke, E. M. (1983).Influence of jojoba meal upon growth and reproductionin mice. In Jojoba and Its Uses Through 1982, Pro-ceedings of the Fifth International Conference on Jo-joba and Its Uses, pp. 93-99. Edited by A. Elias-Cesnik. Tucson, AZ: Office of Arid Lands Studies,University of Arizona.

278. Webster, J., Beck, W. & Ternai, B. (1984). Toxicityand bitterness in Australian Dioscoria bulbifera L. andDioscoria hispida Dennst. from Thailand. J AgricFood Chem 32, 1087-1090.

279. Weissenberg, M. (1989). Calcinogenic glycosides. InToxicants of Plant Origin. Vol. 2, Glycosides, pp. 201-238. Edited by P. R. Cheeke. Boca Raton, FL: CRCPress.

280. Westendorf, J. (1993). Anthranoid derivatives — generaldiscussion. In Adverse Effects of Herbal Drugs. Vol. 2,pp. 105-1 18. Edited by P. A. G. M. De Smet, K. Keller,R. Hansel & R. F. Chandler. Berlin: Springer- Verlag.

281. Westley, J. (1988). Mammalian cyanide detoxificationwith sulphane sulphur. In Cyanide Compounds in Biol-ogy; CIBA Foundation Symposium No. 140, pp. 201-218. Edited by David Evered and Sarah Harnett.Chichester, England: John Wiley & Sons.

282. Wheeler, J. L. & Mulcahy, C. (1989). Consequencesfor animal production of cyanogenesis in sorghum for-age and hay: a review. Trop Grass 23, 193-202.

283. White, W. L. B., Arias-Garzon, D. L, McMahon, J. M.& Sayer, R. T. (1998). Cyanogenesis in cassava: therole of hydroxynitrile lyase in root cyanide production.Plant Physiol 116, 1219-1225.

284. Yannai, S. & Marquardt, R. R. (1985). Induction offavism-like symptoms in the rat: effects of vicine anddivicine in normal and buthionide sulphoxime-treatedrats. JSd Food Agric 36, 1 161-1 168.

285. Yeoh, H. H., Tan, T. K. & Loh, C. M. (1995). Sourcesof fungal linamarases. World J Microbiol Biotechnol11,678-680.

286. Yoshiki, Y., Kudou, S. & Ukubo, K. (1998). Relation-ship between chemical structures and biological activi-ties of triterpenoid saponins from soybean. BiosciBiotechnol Biochem 62, 2291-2299.

287. Zee, J. A., Boudreau, A., Bourgeois, M. & Breton, R.(1988). Chemical composition and nutritional qualityof faba bean (Viciafaba L. Minor) based tofu. J FoodSd 53, 1772-1774.

CHAPTER 5

Chemical Hazards andTheir Control: Toxins

Maurice O. Moss

MICROBIAL TOXINS

Many food commodities are susceptible tocontamination by microorganisms, the subse-quent growth of which can lead to the presenceof toxic compounds in food. Some of the bacte-rial toxins, such as those of Staphylococcusaureus, Clostridium botulinum, and manyGram-negative species associated with food poi-soning are macromolecules such as proteins.However, some bacteria are able to produce lowmolecular weight toxic metabolites, which maycontaminate foods and cause serious poisoning.One example is Burkholderia (Pseudomonas)cocovenenans, which produces toxins such asbonkrekic acid and toxoflavin (see bongkrek poi-soning on page 1 12). Another example is the mac-rocyclic depsipeptide, cereulide, which is recog-nized as the emetic toxin of Bacillus cereus.

Several species of cyanobacteria can also pro-duce very toxic metabolites, but rarely are thesedirectly associated with foods. An exceptionmay be the contamination of species, such asSpirulina, grown as food or as components of"health food" products, with toxigenic species.More directly associated with foods are the tox-ins that are produced by groups of eukaryotic al-gae such as dinoflagellates and diatoms. Thesemay be ingested by shellfish or fish and passthrough the food chain to humans.

A number of species of fungi can producerelatively low molecular weight secondary me-tabolites, which are toxic to humans and domes-ticated animals and are referred to as mycotox-

ins.49 Mycotoxin biosynthesis may be associatedwith the preharvest stage of crop production byfungi that are obligate endophytes of plants,plant pathogens, or members of the flora respon-sible for the decay of senescent plant material.47

However, the highest concentrations of many ofthese toxic metabolites are produced by fungigrowing on postharvest commodities that arestored under inappropriate conditions. 31 Themajority of mycotoxins are especially importantin the context of animal husbandry, but severalare also significant as contaminants of humanfoods and will be dealt with in the following sec-tions. Table 5-1 summarizes the toxins consid-ered, their most common sources, and the foodcommodities most frequently implicated.

A number of fermented foods involve a mold-ripening stage, usually with species of Penicil-lium. Some of these molds are known to be po-tentially toxigenic; thus, strains of P. roquefortican produce PR-toxin but not during the produc-tion of the blue cheeses. Many strains of P.camemberti produce cyclopiazonic acid, andthis compound has been detected in the crusts ofCamembert-type cheese but not in the interior ofthe cheeses.35'36 Cyclopiazonic acid has alsobeen found associated with mold-ripened fer-mented sausage.76 Aspergillus versicolor is acommon member of the surface flora of hardcheeses stored for long periods, and this speciesis known to produce sterigmatocystin, a myc-otoxin that is a biosynthetic precursor of theaflatoxins and is both acutely toxic and carcino-genic but very much less so than aflatoxin Bi.

Table 5-1 Toxic Microbial Metabolites that May Be Associated with Foods

Toxin

Aflatoxin Bi

Ochratoxin A

Patulin

T-2 toxin

Deoxynivalenol

Fumonisin Bi

Cyanoginosin

Bonkrekic acid

Toxoflavin

Saxitoxin

Okadaic acid

Domoic acid

Sources

Aspergillus flavusA. parasiticusA. nomius

Penicillium verrucosumA. ochraceus

P. expansumA. clavatus

Fusarium sporotrichioides

F. graminearum

F. moniliforme

Microcystis aeruginosa

Burkholderia cocovenenans

B. cocovenenans

Alexandrium catenella

Dinophysis fortii

Pseudonitzschia pungens

Commodities

Maize, groundnuts,treenuts, dried figs,spices

Cereals, coffee, spices,dried vine fruits

Apple juice, other fruitjuices, malted barleyresidues

Overwintered cereals

Cereals

Maize

Water

Tempeh bonkrek

Tempeh bonkrek

Mussels

Mussels

Mussels

LDso (mg/kg)

0.5 (dog)9.0 (mouse)

28 (rat)

35 (mouse)

5.2 (rat)

46 (mouse)

?

0.05 (mouse)

6.84 (oral LDiooin mouse)

8.4 (oral in mouse)

0.012 (ip in mouse)

0.2 (LD99

in mouse)

3.6 (mouse)

This mycotoxin can be found in cheeses but usu-ally only at the surface.52 A recent study hasdemonstrated the occurrence of sterigmatocystinin Ras cheese that was purchased in local mar-kets in Egypt.43

AFLATOXINS

Although aflatoxins are produced by a smallnumber of species of the genus Aspergillus , theyare especially widespread because there are es-sentially three routes to the contamination offood commodities.

1. Direct contamination through the moldspoilage of stored products by species suchas A. flavus, A. parasiticus, and A. nomius

2. Preharvest production in the field by theestablishment of an endophytic associa-tion of one of these species with plantssuch as maize and groundnuts, followedby some form of stress on the growingcrop, such as drought

3. Passage through the food chain into ani-mal products, such as milk, following theconsumption by farm animals of contami-nated animal feeds

The aflatoxins are a family of metabolites; themost important of which in terms of both toxic-ity and prevalence is aflatoxin Bi (Figure 5-1).Aflatoxins are acutely toxic, carcinogenic, andimmunosuppressive. Aflatoxins have beenfound in a wide range of tropical and subtropicalproducts such as figs, pistachio and Brazil nuts,spices, peanuts, and maize. Pittet59 provided auseful update on the natural occurrence of afla-toxins and other mycotoxins with detailed infor-mation concerning the incidence and range ofconcentrations found. The most important ofthose commodities that might be used as a rawmaterial for fermented foods is maize, althoughthere have been reports of low concentrations ofaflatoxins in both rice50 and wheat.56 A diverserange of products associated with foods and bev-erages of the Far East, such as rice wine, soysauce, and miso, involves the use of a source ofamylolytic, lipolytic, and proteolytic enzymes.This material, known as koji, traditionally is pro-duced by growing appropriate strains of A. oryzaeon substrates such as rice, wheat, and soya beans.This mold is closely related to A. flavus, and mayindeed be a "domesticated" form of this species.15

Some koji fungi are indistinguishable fromA. flavus but are nontoxigenic. It seems that theselection of strains producing increased levels ofsecreted hydrolytic enzymes has selected strainsthat do not produce aflatoxins.

Aflatoxins are susceptible to both microbialand mammalian metabolism. Indeed, it is a con-

sequence of metabolism in the mammalian liverthat the aflatoxins are toxic (Figure 5-2). Thewide range in both acute and carcinogenic toxic-ity (Table 5-2) in different animal species re-sults from differences in the metabolic activitiesof these animals. Carcinogenicity is associatedwith the formation of aflatoxin epoxide and itssubsequent reaction with guanine residues inDNA, whereas the acute toxicity requires thehydroxylation of this epoxide to form dihydro-xyaflatoxin, which can react with the lysine resi-dues of proteins. The possibility that microor-ganisms could be used to degrade aflatoxins, andhence detoxify contaminated foods, has been re-viewed.5 One of the earliest reports of the suc-cessful removal of aflatoxins by microorgan-isms was that of Lillehoj and his colleagues atthe Northern Regional Research Laboratories inPeoria, Illinois.37'38 Arising from a wide screenof both eukaryotic and prokaryotic microorgan-isms, the bacterium Flavobacterium auranti-acum (NRRL B-184) was the most effective, butthe phenomenon has not been converted into apractical detoxification process.

When cows are fed on feed contaminated withaflatoxin Bi, they secrete a proportion of thecontaminant in their milk as the metabolite, afla-toxin Mi. This compound is less toxic than itsprecursor, but, because it is uniformly distrib-uted in a liquid food such as milk, and veryyoung and elderly people may be exposed to asignificant extent, the European Commission(EC) has set a particularly stringent maximumpermissible level for aflatoxin Mi in milk anddairy products of 0.05 jug/kg. The comparablelevel for aflatoxin Bi in groundnuts, nuts, driedfruit, and cereals for direct human consumptionis 2 jig/kg (EC No 1525/98), although less strin-gent levels are set in other parts of the world.8

Unfortunately, there is no consistent evidencethat aflatoxin Mi is removed during any of themost commonly used processes in the milk in-dustry, such as heat treatment, cold storage, orspray drying.79 Neither is aflatoxin completelydegraded during the fermentation processesused in the manufacture of cheese, cream, orbutter, although it may be partitioned betweenthe different components of each process. Thus,Figure 5-1 Aflatoxin B1.

Aflatoxin B1

Microbialdetoxification

Parasiticol

Aflatoxin M1

MILK

Aflatoxin epoxide

codon 249 of p53 gene

CARCINOGENESIS

binding to p53 gene product

HEPATITIS B INFECTION

Dihydrodihydroxyaflatoxin

Reaction with lysine in key enzymes

ACUTETOXICITY

Figure 5-2 The relationship between metabolism and toxicology of aflatoxin B1.

Table 5-2 Oral Acute LDso Values and TDso Values for Carcinogenesis of Aflatoxin Bi

Animal Species LDso (mg/kg body wt) TDso fag/kg body wt/day)

Rabbit 0.3Cat 0.6Dog 0.5-1.0Pig 0.6Baboon 2.0Rat (male) 5.5 1.3-5.8Rat (female) 17.9 6.9-12.5Macaque monkey 7.8Rhesus monkey 156Cynamolgus monkey 848Mouse 9.0 >5,300Hamster 10.2Humans 5.0* 132t

*Based on epidemiological evidence from cases of acute poisoning31

+Based on an analysis of intake versus incidence data78

cheese may retain as much as 60% of the afla-toxin Mi in the milk that is used in its manufac-ture, whereas cream may retain only approxi-mately 10%, and butter as little as 2%.

Aflatoxin Bi may be detoxified during the fer-mentation of milk because Lactococcus lactis(previously known as Streptococcus lactis) is ableto convert it into aflatoxin B2a and aflatoxicol (Fig-ure 5-3).40'41 In fact, the formation of aflatoxin Biamay simply be the result of the reduced pH; thereis the possibility of it being reconverted to the par-ent compound. However, it has been claimed thatthe formation of aflatoxin Bia is not simply a re-sult of acid-catalyzed hydration of aflatoxin Bi,and also that viability is not a prerequisite for theremoval of aflatoxin Bi by probiotic strains ofLactobacillus rhamnosus.2^21 There have beensearches for other microbial systems for the re-moval of aflatoxins from foods and a possibilityactively being investigated is the use of enzymesfrom a nontoxic, edible species of fungus,Armillaria tabescens, which is valued already inChina for its medicinal properties in alleviating anumber of disorders.39

An alternative strategy to detoxifying afla-toxin directly in foods is the possibility of re-moval from the gastrointestinal tract by pro-

biotic bacteria. Such a study has been carried outusing probiotic strains of Lactobacillus andPro-pionibacterium.2 The results from these investi-gations suggest that such probiotic bacteria havea role in reducing the bioavailability of food-borne carcinogens such as aflatoxin Bi. How-ever, the most effective strategies for limitinghuman exposure to aflatoxins are to avoid con-tamination in the first place or a chemical pro-cess, such as ammoniation, to degrade aflatoxinirreversibly in animal feeds.63

OCHRATOXIN A

Ochratoxin A (Figure 5-4) is produced by P.verrucosum in temperate climates and by a num-ber of Aspergillus species, especially A.ochraceus, in warmer parts of the world.48 Och-ratoxin A is most common in cereals of temper-ate countries, such as barley, oats, rye, andwheat,30'45 but has also been found in maize,32

coffee, cocoa, dried vine fruits, wine,59 andbeer.65 Although ochratoxin A can survive fer-mentation processes, it normally does not sur-vive the malting process used in the productionof beer.33 Its presence in beer at very low con-centrations (ca 0.2 ng/ml) may be due to its pres-

Figure 5-3 (a) Aflatoxin B2a (b) Aflatoxicol.

ence in adjuncts that are used in commercialbeer production. A survey of the occurrence ofochratoxins in a range of wines from the Swissretail market showed them to be present at verylow concentrations and to be more frequent, andat higher concentrations, in red wines from more

southerly regions of Europe.82 It is probable thatcontamination precedes the fermentation stage.These results confirm that ochratoxins are notremoved by an alcoholic fermentation.

Ochratoxin A is relatively thermostable, hav-ing a half life at 100 0C of more than 10 hours in

Figure 5-4 Ochratoxin A.

dry wheat and nearly 2.5 hours in moist wheat.7

Several studies have confirmed that ochratoxinA will survive in most food processes involvinga heating stage. Like aflatoxin, ochratoxin A canalso pass through the food chain and may befound in meat products, especially of the pig,32

but it does not seem to be secreted effectivelyinto cow's milk.75 However, surveys in Scandi-navian countries have shown the occurrence ofochratoxin A in cow's milk in Sweden9 and Nor-way.72 Because of the low biotransfer of ochra-toxin A from animal feeds to milk, Skaug72

speculated that the inhalation of contaminatedairborne particles may be the route into cow'smilk. If ochratoxin is not normally present inmilk, it would not be expected to occur in dairyproducts. However, some cheeses are readilycontaminated with molds, and some of the moldsisolated from cheeses have been shown to produceochratoxin in the laboratory. Scott,64 in his detailedreview of the occurrence of mycotoxins in dairyproducts, concluded that "cheese is generally agood substrate for fungal growth but a poor sub-strate for experimental mycotoxin production."(P22i) There are a few reports of the natural occur-rence of ochratoxin A in moldy cheese, and theseare referenced in this review, but the molds useddeliberately for the production of mold-ripenedcheeses (i.e., P. roqueforti and P. camemberti) donot produce this mycotoxin.

The transfer of ochratoxin from animal feedsto animal tissues, such as muscle, liver, and kid-neys, combined with the extended residencetime for ochratoxin A in animal tissues, leads tothe possibility of its presence in meat products.There are many reports of the occurrence of och-ratoxin A in kidneys, liver, and even sausages,and these have been documented extensively.34

The contamination of meat products by transferfrom animal feeds should be distinguished fromthe occurrence of ochratoxin A in moldy meatproducts such as smoked pork, other smokedmeats, and sausages, in which much higher lev-els of ochratoxin A can be found.34

At the acute level, ochratoxin A is anephrotoxin, which is certainly responsible formost cases of porcine nephropathy and has beensuspected to be an etiological agent in Balkan

endemic nephropathy. A detailed risk assess-ment of this mycotoxin has been carried out34

and it seems prudent to assume that it is also car-cinogenic.17 For this reason, and because there isno doubt regarding human exposure to ochra-toxin A from a range of foods, the member coun-tries of the European Union are presently seek-ing to agree to maximum levels in foods forhuman consumption; these are likely to includemaximum acceptable levels in fermented bever-ages such as beer and wine.

PATULIN

Patulin (Figure 5-5) is produced by a numberof species of Penicillium, Aspergillus, andByssochlamys, but, in the context of humanfoods, the most important species is P.expansum. This mold is associated especiallywith a soft rot of apples, but may also occur on awide range of other fruits. P. expansum is able toform a rapid brown soft rot in apples, once infec-tion has been established, and it becomes imme-diately recognizable once the conidiophoresbearing large numbers of blue green spores formas pustules on the surface of the rot. Althoughinfection is usually through a wound and the rotis usually very evident, some apple varieties canbe infected from within, and, although the applelooks superficially sound, the core may be in-fected and contaminated with patulin. It is un-likely that fresh fruit will be a hazard because assoon as mold contamination is apparent, it willnormally be discarded because of the obviousvisual and taste defects. However, an organolep-

Figure5-5 Patulin.

tically acceptable fruit juice can be expressedfrom fruit containing some rot, and patulin isstable at the low pH values of most fruit juices.

The natural occurrence of patulin in commer-cial apple juice was reported as long ago as 1972.69

Surveys since then have demonstrated that thecontamination of fruit juices with patulin is a con-tinuing nuisance.10'44'61 Patulin was discoveredoriginally as a potentially useful broad spectrumantibiotic, but the acute toxicity precluded its use.The report of Dickens & Jones in 196119 suggestedthat, at high enough doses, patulin could inducesarcomas in experimental animals at the site of in-jection and hence may be carcinogenic. However,several studies since then have failed to provideconclusive evidence that patulin is carcinogenic,26

but the United Kingdom has set an advisory maxi-mum level of 50 ug/1 in apple juice for human con-sumption, and there is not yet any statutory limitfor patulin in the European Community. A reportof the U.K. Ministry of Agriculture, Fisheries andFood44 demonstrated that a few samples of freshapple juice taken from retail outlets in the UnitedKingdom were contaminated with patulin at levelsthat were above the advisory limit. A detailed as-sessment of this report and others, as well as theanalytical difficulties associated with the determi-nation of low concentrations of patulin in applejuice, is available.71

It has been known for some time that patulindisappears during the fermentation of apple juiceto cider using the yeast Saccharomycescerevisiae.23 The microbial decomposition ofpatulin requires the yeast to be viable; it is an in-ducible phenomenon, and occurs during fermenta-tive metabolism rather than respiratory metabo-lism. More recently, it has been shown that patulinis metabolized to a number of products includingascladiol during a yeast fermentation.71 If due dili-gence is paid to the quality of apple juice used inthe manufacture of cider, it seems unlikely thatpatulin will be a problem in cider.

FUSARIUM TOXINS

The genus Fusarium contains several speciesthat are important plant pathogens causing seri-ous losses in agriculture and horticulture. They

are thus of special importance in the field duringthe growth and development of a crop. Somespecies are also capable of continuing growthand metabolic activity postharvest, but they re-quire higher water activities (Aw) than most spe-cies ofPenicillium andAspergillus. Fusarium isassociated with a very wide range of secondarymetabolites, many of which are toxic to farmanimals and humans. The three groups that areparticularly important in human foods are thepolyketide-derived zearalenone, the sesquiter-penoid trichothecenes, and the recently discov-ered fumonisins. Bennett & Richard4 reviewedthe effects of processing on these toxins in con-taminated grains and it is clear that they arestable during wet and dry milling and ethanolfermentations with the exception of deoxyni-valenol, for which there is conflicting evidenceconcerning the effect of fermentation. Althoughthese processes may not destroy the commonfusarial toxins, they do influence their segrega-tion among the various fractions of the process.Thus, for example, wet milling will producetoxin-free starch from maize, but the other prod-ucts, used in animal feeds, will have higher lev-els of zearalenone than the starting material.

Zearalenone

Zearalenone (Figure 5-6) has little or no acutetoxicity but has potent estrogenic activity and isresponsible for a serious disorder in pigs known asvulvovaginitis. Zearalenone is produced by sev-eral species of Fusarium, but the most importantare F. graminearum and F. culmorum, both ofwhich may be responsible for head rot (often char-acterized by a pink or red discoloration of the de-

Figure 5-6 Zearalenone.

veloping grain) in cereals such as maize, wheat,and barley. The ecology of Fusarium in cerealshas been reviewed in detail by Chelkowski,13 andthere are many reports of the occurrence ofzearalenone, with or without other Fusarium tox-ins, in cereals.25' 59> 73 Zearalenone can be formed inthe field (up to 21 mg/kg ),55 but the largest con-centrations are usually associated with poor stor-age postharvest. Concentrations as high as 2.9 g/kg have been reported.46 A survey of Canadianbeers for Fusarium mycotoxins failed to demon-strate the occurrence of zearalenone,66 and itseems unlikely that this metabolite will occur atsignificant levels in fermented milk products.

Trichothecenes

The trichothecenes form a very large familyof sesquiterpene metabolites, the most important

of which, in the context of foods, are producedby species of Fusarium. The terrible outbreaksof alimentary toxic aleukia in humans, and ofhemorrhagic moldy corn toxicosis in farm ani-mals, are especially associated with T-2 toxin(Figure 5-7 a), which is produced primarily by F.sporotrichioides and F. poae. This is one of agroup of very toxic trichothecenes that fortu-nately are relatively rare in crops grown for hu-man consumption. The conditions leading tooutbreaks of alimentary toxic aleukia are hope-fully avoidable (i.e., widespread famine follow-ing a war). A much more common trichotheceneis deoxynivalenol (Figure 5-7b), which isformed by F. graminearum, F. culmorum, andrelated species. Indeed, there are some yearswhen samples from crops such as barley mayshow 100% incidence of deoxynivalenol (DON)because these species of Fusarium are patho-

Figure 5-7 (a) T-2 toxin (b) Deoxynivalenol.

gens that may establish themselves in the field,causing red scab diseases, and continue their ac-tivity postharvest.

During the period 1996-1998, DON waswidespread in cereals throughout the temperateparts of the world, with incidences from 43% to100% and concentrations from 2 jug/kg to62,050 jag/kg.59 There is a clear account of thetype of weather conditions that occurred duringthe growing season for maize in Maryland andDelaware during 1994, which led to DONcontamination prior to harvest.77 Although notas toxic as T-2 toxin, DON, like all thetrichothecenes, is immunosuppressive and bio-logically active at concentrations very muchless than the LD50 of 50-70 mg/kg. DON is astable molecule and may pass into fermentedcereal products, but it does not appear to be aproblem in fermented dairy products. It wouldbe surprising if DON were not found in beersand, indeed, it has been reported that 29 of 50samples of Canadian and imported beer ana-lyzed contained DON.66 Although the majorityof beers had very low concentrations of DON,nine had more than 5 ng/ml, and a single samplehad as much as 50.3 ng/ml.

Fumonisins

The fumonisins are produced by F.moniliforme and related fusaria that do not pro-duce trichothecenes. Fumonisin Bi (Figure 5-8)is known to cause equine encephalomalacia, pul-monary edema in pigs, and a number of other ill-nesses in a range of animal species. It may beassociated with esophageal carcinoma in hu-mans62 and has recently been the subject of amonograph.28 An assessment of human exposureto the fumonisins for people in the Netherlandswith a special emphasis on those people withgluten intolerance who would be especially atrisk has been carried out.18 The fumonisins areassociated with maize and maize products, re-flecting the host specificity of the molds produc-ing it. In those commodities, it is remarkablywidespread and can occur in relatively high con-centrations. In his extensive review, Pittet59 pro-vided documentation of surveys in many countriesduring the period 1995-1998 with incidence ofoccurrence ranging from 20% to 100% and con-centrations from 10 |iig/kg to 37,650 jug/kg.These surveys included maize products such aspolenta, corn flakes, and popcorn.

Figure 5-8 Fumonisin B1.

As with DON, it would seem likely thatfumonisins will be found in beers in whichmaize products are used at some stage during themanufacture, but there is no evidence of trans-mission from animal feeds to milk and hence tomilk products. By using immunoaffmity col-umns for extraction and cleanup, it has beenshown that 86% of samples of beer that was pur-chased in retail outlets in Lincoln, Nebraska,contained fumonisin Bi.24 The Lincoln study fol-lowed an earlier report of fumonisin contamina-tion of beer in Canada that had included beersthat were imported from the United States.67'70

These authors agreed that the most likely sourceof fumonisin in beer is the maize grits that areused as a brewing adjunct.

The fumonisins are relatively stable to el-evated temperatures and survive a range ofcooking, baking, and frying processes. Thus, af-ter baking corn muffins spiked with fumonisinBi for 20 minutes at 175 or 200 0C, as much as84% and 72%, respectively, survived.29 A sig-nificant reduction when spiked corn masa wasfried at 140-170 0C for up to six minutes wasalso found. Even frying contaminated chips for15 minutes at 190 0C only reduced the fumonisinlevel by 67%. However, it has also been reportedthat, during extrusion cooking of corn grits, thereduction of fumonisin Bi depended on severalfactors, including moisture content and whether ornot mixing screws were used.12 One of the prob-lems in assessing the significance of the break-down of the fumonisins is that they are esters of anaminopentol compound, and some food processeslead to the hydrolysis of the ester groups to the par-ent compound, which is still toxic.

The preparation of tortillas in Central andSouth America requires the treatment of maizewith lime to produce nixtamal before cooking.Scott & Lawrence68 validated the methodologyfor measuring the aminopentol formed fromfumonisin Bi in calcium hydroxide-processedfoods such as tortilla and nacho chips, tacoshells, and air-dried corn tortillas. They foundsignificant levels of the aminopentol, but alwaysat lower levels than the fumonisin Bi from whichit was derived. This implies that the latter hadpartly survived the alkaline process. Tortillas

from villages in Guatemala may have as much as185 mg kg'1 of APi (i.e., the aminopentol de-rived from the hydrolysis of fumonisin Bi) andalso still contain up tolO mg kg"1 of fumonisin Biitself.42 APi may not always be formed duringcooking and food processing of contaminatedmaize, even if fumonisin levels are reduced.57

ALTERNARIA TOXINS

Species ofAlternaria cause dry and soft blackrots of a range of commercially important cropsand are often associated with the postharvestspoilage of fruits and vegetables.58 The most com-mon species is A. alternata, which can produce anumber of toxic metabolites, the most importantof which is tenuazonic acid (Figure 5-9). Othertoxins include alternariol and its monomethylether, altenuene, and the altertoxins. The analysisof such a complex mixture requires a multitoxinmethod.51 One or more of these toxins have beenisolated from a number of cereals, oilseed rape,tomatoes, sunflower seed, and olives. There isthe potential for these toxins to occur in theproducts of cereals, oilseeds such as sunflower,fruit juices, and tomato products, but there is noinformation regarding the effects of food pro-cessing on the survival ofAlternaria toxins intofermented foods. Although isolates ofAlterna-ria have been reported as part of the mold floraof some cheeses,64 it is unlikely that Alternariatoxins will have any significance for humanhealth from fermented foods. A detailed accountofAlternaria species and their toxins is given byBottalico & Logrieco.6

Figure 5-9 Tenuazonic acid.

BACTERIAL AND ALGAL TOXINS

The conditions for the growth of most foodpoisoning bacteria are generally well under-stood, and the control of growth can be effectedby heat treatment, refrigeration, reduced Aw andlow or high pH, alone or in combination. One ormore of these parameters can also be combinedwith the presence of both synthetic or naturalgrowth inhibitors. However, once bacterial tox-ins have been formed and released into the food,they are generally resistant to the conditions thatinhibit or kill the producing organism.

Botulinum toxins are relatively largepolypeptides (i.e., molecular weights of approxi-mately 150 kDa) that are usually complexedwith proteins as a progenitor toxin when they arereleased into the food. Such complexes are morestable at elevated temperatures and low pH thanthe pure neurotoxins. Of 14 outbreaks of botu-lism associated with dairy products between1912 and 1997, the majority involved cheese orcheese spreads.3'14'60 An especially serious out-break in the United Kingdom during 1989 in-volving 27 cases and one death was a result ofthe consumption of hazelnut yogurt that wascontaminated with type B neurotoxin producedby a proteolytic strain of C. botulinum.54 Theproblem was not with the yogurt itself, whichhas a pH too low for growth and toxin produc-tion, but with the canned hazelnut conserve thatwas added to it. This was occasioned by achange in production from a product with a highsucrose content receiving a mild heat treatmentto a product in which sucrose was replaced withaspartame, in response to a perceived public de-mand for low sugar products. The heat treatmentremained unchanged and was insufficient to killC. botulinum spores, but the high sucrose con-tent prevented growth in the original formula-tion. The elevated Aw of the new formulation al-lowed growth and toxin formation in thehazelnut conserve and, although the pH of theyogurt to which it was added prevented furthergrowth of the organism, it did not inactivate thetoxin that had already been produced.

The enterotoxins of S. aureus are small singlepolypeptide chain proteins (i.e., molecular

weights 26 to 30 kDa) that are more heat resistantthan the producing organism. Although fermenta-tion processes would normally prevent growth ofthis organism, there have been instances of entero-toxin production in cheese when the starter culturefailed to grow fast enough,80 and in salami sausagewhen a change in production conditions again al-lowed S. aureus to outcompete the organisms usedfor the fermentation.53'74

The emetic toxin of B. cereus known ascereulide is a cyclic dodecadepsipeptide1 that isstable to proteolytic enzymes and remarkablyheat stable (i.e., no loss of activity after 90 min-utes at 121 0C). Cereulide is usually produced incooked rice, pasta, and noodles, and is very un-likely to occur in fermented foods.

Bongkrek Poisoning

An Indonesian food known as tempeh bong-krek is made by inoculating coconut presscakewith the mold Rhizopus oligosporus. Contamina-tion of this food by the pseudomonad E. cocoven-enans, previously known as P. cocovenenans*1

unfortunately is not uncommon and has been re-sponsible for quite a number of deaths. B.cocovenenans produces two toxic metabolites,toxoflavin (Figure 5-10) and bongkrekic acid(Figure 5-11), which has been shown to be aheptenetrioic acid.16 Normally, the rapid growthof R. oligosporus over and into the solid sub-strate effectively inhibits bacterial growth, al-though bonkrekic acid formation is not necessar-ily directly correlated with bacterial numbers.

Figure 5-10 Toxoflavin.

Figure 5-11 Bongkrekic acid.

Outbreaks of bonkrek poisoning seem to oc-cur most often when coconut presscake is mixedwith excessive amounts of coconut milk. Thepresence of both an aqueous environment andelevated levels of lipids seems to encourage bac-terial growth and bongkrekic acid formation. Itseems that temperature, pH, moisture, and saltconcentration themselves do not fully explainwhy some tempeh bonkrek becomes toxic andanother factor, such as the presence of appropri-ate lipids, may be implicated. In a useful reviewof the conditions influencing the formation ofthese toxins, it was specifically demonstratedthat fatty acids have an important role in the for-mation of bongkrekic acid.22 Although compa-rable in acute toxicity, bongkrekic acid is usu-ally present at much higher concentrations thantoxoflavin and is most likely to be of greatestsignificance in causing illness. A few hours afterconsumption of toxic tempeh bongkrek, peoplecomplain of malaise, abdominal pains, dizzi-ness, sweating, and fatigue. Coma and death canfollow within 20 hours of the onset of symp-toms. The edible jelly fungus (Tremellafuciformis), which is cultivated in several coun-tries in southeast Asia, is contaminated fre-quently with B. cocovenenans, and has also beenimplicated in bongkrek poisoning.

Cyanobacterial Toxins

A number of species of freshwatercyanobacteria belonging to the generaMicrocystis, Anabaena, and Aphanizomenoncan form extensive blooms in standing water andmay cause deaths of animals drinking the con-taminated water.11 Cyanoginosin (Figure 5-12),a toxic metabolite of Microcystis aeruginosa, isan hepatotoxin. It is unlikely that these toxinswill find their way into fermented foods, al-though there may be some concern that they maycontaminate samples of Spirulina that are col-lected from the wild and used as food and healthproducts.

Algal Toxins

Among the eukaryotic algae, it is the di-noflagellates and a small group of diatoms thatcause concern because of their ability to producevery potent toxins. A number of toxic responsesto contamination of seafoods occur, of whichparalytic shellfish poisoning, diarrheal shellfishpoisoning, and neurotoxic shellfish poisoningare all associated with marine dinoflagellates,and amnesic shellfish poisoning with species ofmarine diatom. Table 5-3 lists a selection of the

Figure 5-12 Cyanoginosin.

Table 5-3 A Selection of Eukaryote Algal Toxicoses

Toxicosis

Paralytic shellfishpoisoning (PSP)

Diarrheal shellfishpoisoning (DSP)

Neurotoxic shellfishpoisoning (NSP)

Ciguaterapoisoning

Amnesic shellfishpoisoning (ASP)

Associated Species

DinoflagellatesAlexandrium catanella

Dinophysis fortii

Gymnodium breve

Gambierdiscus toxicus

DiatomsPseudonitzschia pungens

Toxins

Saxitoxin (Figure 5-13)

Okadaic acid (Figure 5-14)

Brevitoxin

Ciguatoxin

Domoic acid (Figure 5-15)

Figure 5-13 Saxitoxin. Figure 5-15 Domoic acid.

Figure 5-14 Okadaic acid.

species and toxins involved27 (Figure 5-13, 5-14, and 5-15). These toxins are generally resis-tant to the temperatures involved in cooking andare likely to be unaffected by the salting and

REFERENCES

1. Agata, N., Ohta, M., Mori, M. & Isobe, M. (1995). Anovel dodecadepsipeptide, cereulide, is an emetic toxinof Bacillus cereus. FEMS Microbiol Lett 129, 17-20.

2. Ahokas, J., El-Nezami, H., Kankaanpaa, P., Mykkanen,H. & Salminen, S. (1998). A pilot clinical study examin-ing the ability of a mixture of Lactobacillus and Propi-onibacterium to remove aflatoxin from the gastrointesti-nal tract of healthy Egyptian volunteers. Revue Med Vet149, 568.

3. Aureli, P., Franciosa, G. & Pourshaban, M. (1996).Foodborne botulism in Italy. Lancet 348, 1594.

4. Bennett, G. A. & Richard, J. L. (1996, May). Influenceof processing onFusarium mycotoxins in contaminatedgrains. Food Tech, 235-238.

acidification that is associated with fermentedfish products. In any case, it seems unlikely thatthe species of shellfish and fish involved will beused in the production of any fermented foods.

5. Bhatnagar, D., Lillehoj, E. B. & Bennett, J. (1991). Bio-logical detoxification of mycotoxins. In Mycotoxins andAnimal Foods, pp. 815-826. Edited by J. E. Smith & R.S. Henderson. Boca Raton, FL: CRC Press.

6. Bottalico, A. & Logrieco, A. (1998). Toxigenic Alterna-ria species of economic importance. In Mycotoxins inAgriculture and Food Safety, pp. 65-108. Edited by K.K. Sinha & D. Bhatnagar. New York: Marcel Dekker.

7. Boudra, H., Le Bars, P. & Le Bars, J. (1995). Thermo-stability of ochratoxin A in wheat under two moistureconditions. Appl and Environ Microbiol 61,11 56-1 158.

8. Boutrif, E. & Canet, C. (1998). Mycotoxin preventionand control: FAO programmes. Revue Med Vet 149,681-694.

9. Breitholtz-Emmanuelsson, A., Olsen, M., Oskarsson,A., Palminger, I. & HuIt, K. (1993). Ochratoxin A incow's milk and in human milk with corresponding hu-man blood samples. J AOAC Ml 76, 842-846.

10. Burda, K. (1992). Incidence of patulin in apple, pear,and mixed fruit products marketed in New South Wales.J Food Prot 55, 796-798.

11. Carmichael, W. W. (1992). Cyanobacterial secondarymetabolites: the cyanotoxins. J Appl Bacterial 72, 445-459.

12. Castelo, M. M., Katta, S. K., Sumner, S. S., Hanna, M.A. & Bullerman, L. B. (1998). Extrusion cooking re-duces recoverability of fumonisin Bi from extruded corngrits. J Food Sd 63, 696-698.

13. Chelkowski, J. (1998). Distribution ofFusarium speciesand their mycotoxins in cereal grains. In Mycotoxins inAgriculture and Food Safety, pp. 45-64. Edited by K. K.Sinha & D. Bhatnagar. New York: Marcel Dekker.

14. Collins-Thompson, D. L. & Wood, D. S. (1993). Con-trol in dairy products. In Clostridium Botulinum. Ecol-ogy and Control in Foods, pp. 261-277. Edited by A. H.W. Hauschild & K. L. Dodds. New York: MarcelDekker.

15. Cruickshank, R. H. & Pitt, J. I. (1990). Isoenzyme pat-terns mAspergillusflavus and closely related species. InModern Concepts in Penicillium and Aspergillus Clas-sification, pp. 259-265. Edited by R. A. Samson & J. I.Pitt. New York: Plenum.

16. De Bruijn, J., Frost, D. J., Nugteren, D. H., Gaudemer,A., Lijmbach, G. W. M., Cox, H. C. & Berends, W.(1973). The structure of bongkrekic acid. Tetrahedron29, 1541-1547.

17. De Groene, E. M., Jahn, A., Horbach, G. J. & Fink-Gemmels, J. (1996). Mutagenicity and genotoxicity ofthe mycotoxin ochratoxin A. Environ Toxicol andPharmacol 1,2 1-26.

18. De Nijs, M., Van Egmond, H., Nauta, M., Rombouts, F.M. & Notermans, S. H. W. (1998). Assessment of humanexposure to fumonisin Bl. J Food Prot 61, 879-884.

19. Dickens, F. & Jones, H. E. H. (1961). Carcinogenic ac-tivity of a series of reactive lactones and related sub-stances. Br J Cancer 15, 85-100.

20. El-Nezami, H., Kankaanpaa, P., Salminen, S. &Ahokas, J. (1998). Ability of dairy strains of lactic acidbacteria to bind a common food carcinogen, aflatoxinBi. Food and Chem Toxicol 36, 321-326.

21. El-Nezami, H., Kankaanpaa, P., Salminen, S. &Ahokas, J. (1998). Physicochemical alterations enhancethe ability of dairy strains of lactic acid bacteria to re-move aflatoxin from contaminated media. / Food Prot61,466^68.

22. Garcia, R. A., Hotchkiss, J. H. & Steinkraus, K. H.(1999). The effect of lipids on bongkrekic (Bongkrek)acid toxin production by Burkholderia cocovenenans incoconut media. Food Add and Contam 16, 63—69.

23. Harwig, J., Scott, P. M., Kennedy, B. P. C. & Chen, Y. -K. (1973). Disappearance of patulin from apple juicefermented by Saccharomyces sp. J Inst ofTechnol andAlimentation 7, 295-312.

24. Hlywka, J. J. & Bullerman, L. B. (1999). Occurrence offumonisins Bi and 62 in beer. Food Add and Contam 16,319-324.

25. Hussein, H. M., Franich, R. A., Baxter, M. & Andrew, I.G. (1989). Naturally occurring Fusarium toxins in NewZealand maize. Food Add and Contam 6, 49-58.

26. International Agency for Research on Cancer. (1986).Patulin. InIARC Monographs on the Evaluation of Carci-nogenic Risks to Humans. Some Naturally Occurring andSynthetic Food Components, Furocoumarins and UltraViolet Reductions. (Lyon, France: IARC) 40, 83-98.

27. International Commission for Microbiological Specifi-cations in Foods. (1996). Seafood toxins of microbio-logical origin. In Microorganisms in Foods, pp. 266-279. London: Blackie Academic and Professional.

28. Jackson, L. S., DeVries, J. W. & Bullerman, L. B., eds.(1996). Fumonisins in Food. New York: Plenum Press.

29. Jackson, L. S., Katta, S. K., Fingerhut, D. D., DeVries, J.W. & Bullerman, L. B. (1997). Effects of baking andfrying on the fumonisin Bi content of corn-based foods.JAgric and Food Chem 45, 4800^805.

30. J0rgensen, K., Rasmussen, G. & Thorup, I. (1996). Och-ratoxin A in Danish cereals 1986-1992 and daily intakeby the Danish population. Food Add and Contam 13,95-104.

3 1 . Krishnamachari, K. A. V. R., Bhat, R. V., Nagarajon, V.& Tilek, T. B. G. (1975). Hepatitis due to aflatoxicosis.Lancet i, 1061-1063.

32. Krogh, P. (1987). Ochratoxins in food. In Mycotoxins inFood, pp. 97-121. Edited by P. Krogh. London: Aca-demic Press.

33. Krogh, P., HaId, B., Gjertsen, P. & Myken, F. (1974).Fate of ochratoxin A and citrinin during malting andbrewing experiments. Appl Microbiol 28, 31-34.

34. Kuiper-Goodman, T. & Scott, P. M. (1989). Risk assess-ment of the mycotoxin ochratoxin A. Biomed andEnviron Sd 2, 179-248.

35. Le Bars, J. (1979). Cyclopiazonic acid production byPenicillium camemberti Thorn., and natural occurrenceof this toxin in cheese. Appl and Environ Microbiol 36,1052-1055.

36. Le Bars, J. (1990). Detection and occurrence ofcyclopiazonic acid in cheeses. J Environ Pathol, ToxicolandOncol 10, 136-137.

37. Lillehoj, E. B., Ciegler, A. & Hall, H. H. (1967). Afla-toxin Bi uptake by Flavobacterium aurantiacum and re-sulting toxic effects. JBacteriol 93, 464-^71.

38. Lillehoj, E. B., Stubblefield, R. D., Shanon, G. M. &Shotwell, O. L. (1971). Aflatoxin Mi removal fromaqueous solutions by Flavobacterium aurantiacum.Mycopath Mycol Appl 45, 259-266.

39. Liu, D. -L., Yao, D. -S., Liang, R., Ma, L., Cheng, W. -Q. & Gu, L. -Q. (1998). Detoxification of aflatoxin Biby enzymes isolated fromArmillariella tabescens. Foodand Chem Toxicol 36, 563-574.

40. Megalla, S. E. & Hafez, A. H. (1982). Detoxification ofaflatoxin Bi by acidogenous yoghurt. Mycopathologia77,89-91.

41. Megall, S. E. & Mohran, M. A. (1984). Fate of aflatoxinBi in fermented dairy products. Mycopathologia 88, 27-29.

42. Meredith, F. L, Torres, O. R., De Tejada, S. S., Riley, R.T. & Merrill, A. H. (1999). Fumonisin Bi and hydro-lyzed fumonisin Bi (APi) in tortillas and nixtamalizedcorn (Zea mays L.) from two different geographical lo-cations in Guatemala. JFoodProt 62, 1218-1222.

43. Metwally, M. M., El-Sayed, A. M. A., Mehriz, A. M. &Abu Sree, Y. H. (1997). Sterigmatocystin: incidence,fate and production by A. versicolor in Ras cheese. Myc-otoxin Res 13, 61-66.

44. Ministry of Agriculture, Fisheries and Food. (1993).Mycotoxins: Third Report. The Thirty Sixth Report ofthe Steering Group on Chemical Aspects of Food Sur-veillance Sub-Group on Mycotoxins. Food SurveillancePaper 'No. 36. London: HMSO.

45. Ministry of Agriculture, Fisheries and Food. (1995).Surveillance of UK cereals for ochratoxin A. Ministry ofAgriculture Fisheries and Food. Food Surveillance In-formation Sheet No. 48. London: HMSO.

46. Mirocha, C. J. & Christensen, C. M. (1974). Oestrogenicmycotoxins synthesised by Fusarium. In Mycotoxins,pp. 129-148. Edited by I. F. H. Purchase. Amsterdam:Elsevier.

47. Moss, M. O. (1996). Mycotoxins. Mycolog Res 100,513-523.

48. Moss, M. O. (1996). Mode of formation of ochratoxinA. Food Add and Contam 13 (Suppl.), 5-9.

49. Moss, M. O. (1998). Recent studies of mycotoxins. JAppl Microbiol 84 (Suppl.), 62S-76S.

50. Miihlemann, M., Liithy, J. & Hiibner, P. (1997). Myc-otoxin contamination of food in Equador. A: aflatoxins.Mitt Gebiete Lebensm Hyg 88, 474-496.

51. Nawaz, S., Scudamore, K. A. & Rainbird, S. C. (1997).Mycotoxins in ingredients of animal feeding stuffs: I.determination ofAlternaria mycotoxins in oilseed rapemeal and sunflower seed meal. Food Add and Contam14, 249-262.

52. Northolt, M. D., Van Egmond, H. P., Soentoro, P. S. &Deyll, W. E. (1980). Fungal growth and the presence ofSterigmatocystin in hard cheese. J Association of Offi-cial Analytical Chemists 63, 115-119.

53. Nychas, G. J. E. & Arkoudelos, J. S. (1990). Staphylo-cocci: their role in fermented sausages. J Appl Bacteriol69 (Suppl.), 167S-188S.

54. O'Mahony, M., Mitchell, E., Gilbert, R. J., Hutchinson,D. N., Begg, N. T., Rodhouse, J. C. & Morris, J. E.(1990). An outbreak of foodborne botulism associatedwith contaminated hazelnut yoghurt. Epidemiol and In-fect 104, 389-395.

55. Park, J. J., Smalley, E. B. & Chu, F. S. (1996). Naturaloccurrence of Fusarium mycotoxins in field samplesfrom the 1992 Wisconsin corn crop. Appl and EnvironMicrobiol 62, 1642-1648.

56. Pineiro, M. S., Dawson, R. & Costarrica, M. L. (1996).Monitoring programme for mycotoxin contamination inUruguayan food and feeds. Nat Tox 4, 242-245.

57. Pineiro, M. S., Miller, J., Silva, G. & Musser, S. (1999).Effect of commercial processing on fumonisin concen-trations of maize-based foods. Mycotoxin Res 15, 2-12.

58. Pitt, J. I. & Hocking, A. D. (1997). Fungi and FoodSpoilage. 2nd edn. London: Blackie Academic and Pro-fessional.

59. Pirtet, A. (1998). Natural occurrence of mycotoxins infoods and feeds — an updated review. Revue Med Vet149, 479^92.

60. Pourshafie, M. R., Saifie, M., Shafiee, A., Vahdani, P.,Aslani, M. & Salemian, J. (1998). An outbreak of food-borne botulism associated with contaminated locallymade cheese in Iran. Scandinavian J Infect Dis 30, 92-94.

61. Preita, J., Moreno, M. A., Diaz, S., Suarez, G. &Dominguez, L. (1994). Survey of patulin in apple juiceand childrens' apple food by the diphasic dialysis mem-brane procedure. J Agric and Food Chem 42, 1701-1703.

62. Rheeder, J. P., Marasas, W. F. O., Thiel, P. G.,Sydenham, E. W., Shephard, G. S. & Van Schalkwyk,D. J. (1992). Fusarium moniliforme and fumonisins incorn in relation to human esophageal cancer in Transkei.Phytopathology 82, 353-357.

63. Riley, R. T. & Norred, W. P. (1999). Mycotoxin preven-tion and decontamination — a case study on maize.Food, Nutri and Agric 23, 25-32.

64. Scott, P. M. (1989). Mycotoxigenic fungal contaminantsof cheese and other dairy products. In Mycotoxins inDairy Products, pp. 193-259. Edited by H. P. VanEgmond. London: Elsevier Applied Science.

65. Scott, P. M. & Kanhere, S. R. (1995). Determination ofochratoxin A in beer. Food Add and Contam 12, 591-598.

66. Scott, P. M., Kanhere, S. R. & Weber, D. (1993). Analy-sis of Canadian and imported beers for Fusarium myc-otoxins by gas chromatography: mass spectrometry.Food Add and Contam 10, 381-389.

67. Scott, P. M. & Lawrence, G. A. (1995). Analysis of beerfor fumonisins. JFoodProt 58, 1379-1382.

68. Scott, P. M. & Lawrence, G. A. (1996). Determinationof hydrolysed fumonisin Bi in alkali processed cornfoods. Food Add and Contam 13, 823-832.

69. Scott, P. M., Miles, M. F., Toft, P. & Dube, J. G. (1972).Occurrence of patulin in apple juice. J Agric and FoodChem 20, 450-451.

70. Scott, P. M., Yeung, J. M., Lawrence, G. A. & Prelusky,D. B. (1997). Evaluation of enzyme linked immuno-sorbent assay for analysis of beer for fumonisins. FoodAdd and Contam 14, 445-450.

71. Scudamore, K. A. (1999). Mycotoxins, an IndependentAssessment of MAFF-Funded Applied Research andSurveillance 1993-1996. Report No. PB 4045. London:Ministry of Agriculture Fisheries and Food (MAFF).

72. Skaug, A. S. (1999). Analysis of Norwegian milk andinfant formulas for ochratoxin A. Food Add and Contam16,75-78.

73. Sohn, H. -B., Seo, J. -A. & Lee, Y. -W. (1999). Co-oc-currence of Fusarium mycotoxins in mouldy andhealthy corn from Korea. Food Add and Contam 16,153-158.

74. U.S. Department of Health, Education and Welfare.(1971). Staphylococcal gastroenteritis associated withsalami. MMWR 20, 253-258.

75. Valenta, H. & GoIl, M. (1996). Determination of ochra-toxin A in regional samples of cow's milk from Ger-many. Food Add and Contam 13, 669-676.

76. Van Rensburg, S. J. (1984). Subacute toxicity of themycotoxin cyclopiazonic acid. Food and Chem Toxicol22, 993-998.

77. Wetter, M. T., Trucksess, M. W., Roach, J. A. & Bean,G. A. (1999). Occurrence and distribution of Fusariumgraminearum and deoxynivalenol in sweet corn ears.Food Add and Contam 16, 119-124.

78. Wogan, G. N. (1992). Aflatoxins as risk factors forhepatocellular carcinoma in humans. Cancer Res 52(Suppl.),2114S-2118S.

79. Yousef, A. E. & Marth, E. H. (1989). Stability and deg-radation of aflatoxin Mi. In Mycotoxins in Dairy Prod-ucts, pp. 127-161. Edited by H. P. Van Egmond. Lon-don: Elsevier Applied Science.

80. Zehren, V. L. & Zehren, V. F. (1968). Examination oflarge quantities of cheese for Staphylococcal enterotoxinA. J Dairy Sd 51, 635-644.

81. Zhao, N. X., Qu, C. F., Wang, E. & Chen, W. X. (1995).Phylogenetic evidence for the transfer of Pseudomonascocovenenans to the genus Burkholderia asBurkholderia cocovenenans comb. nov. Intl J SysBacteriol 45, 600-603.

82. Zimmerli, B. & Dick, R. (1996). Ochratoxin A in tablewine and grape-juice: occurrence and risk assessment.Food Add and Contam 13, 655-668.

CHAPTER 6

Toxic Nitrogen Compounds Producedduring Processing: Biogenic Amines,

Ethyl Carbamides, NitrosaminesM. Hortensia Silla-Santos

INTRODUCTION

Nitrogen compounds are very important in thenutrition of microorganisms, plants, animals,and humans. On the other hand, nitrogen is alsoa component of potentially harmful compoundssuch as amines, amides, andnitrosamines. Thesecompounds can be present in foods as a result offood components and process conditions.

BIOGENIC AMINES

Biogenic amines are basic nitrogenous com-pounds that are formed mainly by the decar-boxylation of amino acids or by amination andtransamination of aldehydes and ketones. Theyare organic bases with low molecular weight andare synthesized by microbial, vegetable, and ani-mal metabolism. Biogenic amines in food andbeverages are formed by the enzymes of the rawmaterial or by microbial amino acid decarboxy-lase activity,7 but it has been found that some ofthe aliphatic amines can be formed in vivo by theamination of corresponding aldehydes.66 Thechemical structure of biogenic amines (Figure6-1) can either be aliphatic (e.g., putrescine, ca-daverine, spermine, spermidine, agmatine), aro-matic (e.g., tyramine, phenylethylamine), or het-erocyclic (e.g., histamine, tryptamine). Aminessuch as polyamines (i.e., putrescine, spermidine,spermine, and cadaverine) are indispensablecomponents of living cells and are important inthe regulation of nucleic acid function and pro-

tein synthesis, and probably also in the stabiliza-tion of membranes.10

Formation of Biogenic Amines in Food

In most foods and beverages, amine formationis generated by the decarboxylation of the corre-sponding amino acid. The precursors of the mainbiogenic amines involved in food poisoning areshown in Figure 6—1. Prerequisites for biogenicamine formation by microorganisms are the(1) availability of free amino acids, (2) presenceof decarboxylase-positive microorganisms, and(3) conditions that allow bacterial growth anddecarboxylase synthesis and activity.

Because amines are formed by enzymatic ac-tivity of the food or bacterial flora, the inhibitionof such activity or the prevention of bacterialgrowth would be very important in order tominimize the amine content of food. One ap-proach to control the bacterial flora is to use pre-servatives. Some spices and herbs have been re-ported to possess antimicrobial activity againstfood spoilage bacteria. Their use, then, could beconsidered provided they are compatible withthe food. Cloves and cinnamon are reported tobe inhibitory to bacterial growth and biogenicamine production by bacteria, but with only aslight effect. Bacteria differ in their sensitivity tospices, and this sensitivity also depends on otherfactors such as temperature.138 Manufacturingconditions influence the production of biogenicamines. Thus, tyramine, putrescine, and cadav-erine concentrations in tempeh can be low or

DecarboxilationAMINO ACID AMINE

HISTAMINEHISTIDINE

TYROSINE TYRAMINE

LYSINE CADAVERINE

ORNITHINE PUTRESCINEcontinues

Figure 6-1 Principal biogenic amines in fermented food and their precursors: (a) histamine and tyramine; (b)cadaverine and putrescine; (c) serotonine and tryptamine; (d) spermine and spermidine.

high depending on the manufacturing process(e.g., soaked soybeans, type of fermentative mi-croorganisms, boiling, home cooking by stew-ing or frying in oil, and storage temperature).86

Histamine production in cheese has been re-lated to factors such as the availability of sub-strate, pH, salt concentration, and temperature(Table 6-1). The proper storage temperature isprobably the most important method of preven-tion.121 Antibiotics such as penicillin and tetra-cycline added to scombroid fish repress hista-mine synthesis,135 though this is not an approachthat is compatible with food use. Amino acid de-carboxylase activity is stronger in an acidic envi-ronment,19'77 and bacteria tend to produce theseenzymes under acid stress conditions as a part of

their defense mechanisms.19'37'126 Several factorsinfluence the pH decrease in fermented foods,such as starter culture, contaminating flora, tem-perature/time, and the use of additives, such asglucono-8-lactone (GDL). The presence of fer-mentable carbohydrates such as glucose en-hances both bacterial growth and their aminoacid decarboxylase activity. Histamine forma-tion was inhibited by salting, with the inhibitionbeing proportional to the brine concentration.113

There are different opinions as to the influenceof time and temperature of storage on the synthesisof biogenic amines. The content of biogenicamines in food seems to increase with both timeand temperature.1 However, it has been reported103

that histamine levels in raw and cooked ground

Figure 6-1 continued

5 - HYDROXYTRYPTOPHAN SEROTONINE

TRYPTOPHAN TRYPTAMINE

ARGININE SPERMIDINE

SPERMINE

beef were unaffected by storage conditions (4, 7,and 10 0C during 12 days), but when the storagetemperature was increased (21 0C), increasedamine concentration was observed.2 Amine con-centrations were also unaffected by cooking, withthe exception of spermine, which decreased dur-ing heat treatment. Wendakoon & Sakaguchi138

also indicated that histamine is thermally stableduring the cooking process.

Oxygen availability also has a significant effecton the biosynthesis of amines. Facultative anaero-bic microorganisms produce less biogenic aminesin anaerobic conditions as compared with aerobicconditions.37 Feedback repression of histidine-de-carboxylase has also been detected when theamount of histamine increases.135

Microorganisms Producing Biogenic Amines

Among the bacteria that are capable of syn-thesizing biogenic amines, many different spe-

cies within a group of related genera may pos-sess a specific decarboxylase. Also, largeinterspecies variations may occur within one ge-nus. Numerous enteric bacteria have been re-ported to possess histidine decarboxylase activ-ity, but only a few of the bacteria found in foodshave been implicated in the formation of toxico-logically significant levels of histamine. Entericbacteria, specifically Morganella morganii(Proteus morganii\ certain strains ofKlebsiellapneumoniae, and a few strains of Hafnia alvei,are prolific histamine producers and are impor-tant in the hygiene offish products.99'133'138 Like-wise, other bacteria have been related to hista-mine formation in fish, including P. vulgaris,Escherichia, Clostridium, Salmonella, and Shi-gella.41 Staphylococcus spp., Vibrio, Pseudomo-nas, and Bacillus spp. have been identified ashistamine-producing bacteria in fermentedfish,99'138'140 as has cadaverine and putrescine

Table 6-1 Factors Affecting Synthesis of Biogenic Amines (BA)

Factors

SubstrateMicroorganisms with

necessary enzyme

PHTemperatureAtmosphere

AntibioticsFermented carbohydratesSpicesOxidase enzymes

Principal Activity

Free amino acids (increase BA)Amino acid decarboxylase (increases BA)

Acid (increases BA)Refrigeration (decreases BA)Aerobiosis (decreased BA in anaerobic

and increased in aerobic conditions)Inhibition of microorganisms (decreases BA)Acididification (increases BA)Inhibition of microorganisms (decreases BA)MAO, DAO (decrease BA)

References

68,83, 1181,2,18,19,37,40,59

17,59, 125, 126, 12717,59,83,115

17,37,5913511313866,67

synthesis capability by M. morganii, Entero-bacter cloacae, Citrobacter freundii, and Serra-tia liquefaciens ."

In cheeses, the lactic flora is dominant duringthe ripening process. Streptococcus faecalis(Enterococcus faecalis)41 has been associatedwith tyramine in cheese23 and in other milk prod-ucts.35 In miso, tyrosine decarboxylase-produc-ing bacteria have been identified as E. faeciumand Lactobacillus bulgaricus, and histamine de-carboxylase has been associated with Lactoba-cillus spp. and L. sanfrancisco.5l>52 Fermentationmay be important in the formation of biogenicamines through the action of added lactic acidcultures or the natural microflora. Amino aciddecarboxylase activity has been shown to de-pend on the composition of the medium and thegrowth phase of the microorganisms, with thehighest amino acid decarboxylase enzyme ac-tivities being detected in the stationary phase.37

The presence of high concentrations of freeamino acids in meat and meat products that areexposed to microbial degradation enhances thepossible formation of large amounts of biogenicamines.109 Amine-producing lactic bacteria suchas L. brevis, L. buchnerii, L. curvatus, L. carnis,L. diver gens (Carnobacterium divergens)41 andL hilgardii have been isolated from meat andmeat products.73'76'77'110'114'122 In salami, hista-mine synthesis has been associated with Gram-

negative bacteria (P. fluorescens, C. freundii,and Acinetobacter calcoaceticus var. anitra-tum), but also with Gram-positive strains, in-cluding micrococci and staphylococci.127

The malo-lactic fermentation bacteriumPediococcus cerevisiae, which withstands thepresence of sulphur dioxide better thanOenococcus oeni, can be regarded as the maincause of histamine formation in wine.12'69 Vari-ous histaminogenic abilities of the yeasts havebeen confirmed in fermentation tests. There isevidence that amino acid decarboxylase en-zymes are not well distributed among the yeastsand the most commonly used malo-lactic bacte-ria. It seems that amino acid decarboxylase ismore frequently found in the groups of entero-bacteria and lactic acid bacteria (LAB). WhenLAB proliferated in beer, the formation of bio-genic amines could be detected.26

Biogenic Amines in Fermented Food

Biogenic amines can be expected in virtuallyall foods that contain proteins or free amino ac-ids and that are subject to conditions enablingmicrobial or biochemical activity. The totalamount of the different amines formed dependsstrongly on the nature of the food and the micro-organisms present.18 Biogenic amines arepresent in a wide range of food products includ-

ing fish products, meat products, dairy products,wine, beer, vegetables, fruits, nuts, and choco-late110 (Table 6-2). In nonfermented foods, thepresence of biogenic amines above a certainlevel is considered as indicative of undesiredmicrobial activity.37 Therefore, the amine levelcould possibly be used as an indicator of micro-bial spoilage, although the presence of biogenicamines does not necessarily correlate with thegrowth of a spoilage flora because not all spoil-ers are decarboxylase positive. Levels of hista-mine, putrescine, and cadaverine usually in-crease during spoilage offish and meat, whereaslevels of spermine and spermidine decrease.18

Scombroid fish have been associated most com-monly with incidents of histamine intoxication(scombrotoxicosis). The formation of histaminein scombroid and other marine fish containingabundant endogenous histidine has been attrib-uted to microbial action rather than to endog-enous histidine decarboxylase activity.37'134 Dur-

ing the preparation of fermented food, one canexpect the presence of many kinds of microor-ganisms, some of them capable of producingbiogenic amines. Most products in which LABgrow contain considerable amounts of pu-trescine, cadaverine, histamine, andtyramine.18

Fish Products

The levels of amines can vary extensively infish products. Trace quantities of putrescine,tyramine, agmatine, and tryptamine have beendetected in Ghananian fermented fish.139 Orni-thine and citrulline were detected as decomposi-tion products of arginine in fish sauce, and hista-mine was also confirmed as a decompositioncompound from histidine, but only in traceamounts.79 In anchovies, a high level of biogenicamines during the manufacturing process can bedue to the long processing time, which favorsbacterial growth.99

Table 6-2 Biogenic Amines Content in Some Foods (ppm)

Food

Fish and fish productsSailfishFermented fish pasteAnchovies

CheesesCheddarSwiss

Meat productsSalchichonFuetDry fermented sausages

Fermented vegetablesSauerkrautSoy sauce

Beers: top-fermentedAleKriek

Beers: botton-fermentedLagerPilsnerNonalcoholicWine

Histamine

1680.0640.0

12.6

1300.02500.0

7.32.2

286.0

10.02740.0

0.65.6

0.71.00.63.3

Tyramine

376.021.6

700.0

280.5190.7

1500.0

20.04660.0

5.022.5

4.95.66.2

Cadaverine

35.038.3

490.0

11.718.9

25.0

0.96.3

0.82.01.0

Putrescine

145.0

7.6

330.0

5.571.0

396.0

50.0

5.74.5

4.15.13.13.3

Reference

47121134

121121

135135121

110121

5555

55555569

A fishy odor is derived from a variety of com-ponents, of which trimethylamine (TMA) ispredominant. TMA and its precursor, TMA N-oxide, are used as a nitrogen source by As-pergillus oryzae.63 Possibly, TMA-using moldscould be applied to remove TMA from relevantfood material.

Cheese and Dairy Products

After fish, cheese is the most commonly im-plicated food associated with histamine poison-ing. The first reported case of histamine poison-ing occurred in 1967 in the Netherlands andinvolved Gouda cheese.121 Many studies havebeen undertaken to determine the amine contentof cheese products, and a variety of amines, suchas histamine, tyramine, cadaverine, putrescine,tryptamine, and phenylethylamine, have beenfound in different cheeses.2'13'23'25'64'80'92'104'132

Meat Products

Biogenic amines can be found in fermentedmeat products as a consequence of microbial ac-tivity related to fermentation, but they may alsobe found in products made from poor qualityraw materials through microbial contamination.Good manufacturing processes for meat prod-ucts, using amino acid decarboxylase negativestarter cultures, results in the formation of verysmall amounts of biogenic amines.28'74 Un-cooked and ripened meats showed higheramounts of histamine and tyramine than cookedmeat products. Maijala et al.14 also detected in-creased concentrations of histamine andtyramine during sausage fermentation. Aminescan also be found after fermentation during stor-age.29 Bauer et al.n reported that the addition ofstarter culture did not affect their formation.Roig-Sagues & Eerola100 observed that the influ-ence of starters on biogenic amine formation inminced meat depended on the kind of decar-boxylating microorganisms that were present inthe raw material. Dry sausages without startermicroorganisms showed variable concentrationsof biogenic amines. This could be due to varia-tions in the ripening time22 and differences in thenatural microflora responsible for fermenta-tion.80'100 Polyamines, putrescine, cadaverine,

histamine, tyramine, and 2-phenylethyl-amine,have also been detected in fermented sausagesby Straub et al. 123 Cured meat products such assausages are frequently implicated in amine poi-soning due to histamine and tyramine.19'73'109'126

Vegetables

A succession of microorganisms is involvedin sauerkraut production. And consequently,biogenic amines, especially putrescine, can beexpected in the brine.18 Other fermented veg-etables where amines have been detected aregreen table olives, cucumbers, and lupin.42

Very low levels of biogenic amines were de-tected in a study of Asian foods. Low levels oftyramine were found in commercial samples ofJapanese pickled vegetables (urume-zuke) and inhim chee (kimchi), a traditional Korean fer-mented cabbage.121 In miso, tyramine and hista-mine have been determined by Ibe et a/.,48"50 butit seems that these amines are not generated inhigh quantities in fermented vegetables.

Wine and Beer

Agmatine, cadaverine, ethanolamine, hista-mine, putrescine, and tyramine are produced inhigh quantities during alcoholic fermentation.20

Many kinds of biogenic amines have been de-tected in both white and red wine: tyramine, his-tamine, tryptamine, monomethylamine, 2-phenethylamine, putrescine, cadaverine,spermidine, iso- and n-amylamine, pyrrolidine,iso- and n-butylamine, iso- and n-propylamine,and ethylamine.51'65'94 Putrescine, histamine, me-thylamine, and tyramine also developed duringmalo-lactic fermentation and the wine agingprocess.12

In beers, the presence of tyramine37 and hista-mine27'54 has been observed. Biogenic amines areformed by the barley enzymes during malting,even under sterile conditions. The grain micro-flora and pitching yeast contaminants seemed tobe responsible for elevated amine levels, espe-cially histamine, in beer. Malt and hops contributeto the amine content of wort and beer. Spermineand spermidine levels decreased sharply duringmashing, whereas the other amines increased,with the exception of putrescine.

Health Effects

Putrescine, spermidine, and spermine are in-dispensable components of all living cells.Polyamines are very stable compounds that areable to resist heat and survive acidic and alkalineconditions. They can interact nonspecificallywith negatively charged structures and, in thesenonspecific roles, they can be replaced by metalions, most often by Mg2+ or Ca2+. Because of theirstructure, they can fulfill a wide range of specificfunctions in cells, such as the control and inhibi-tion of mRNA translation to proteins and regula-tion of the fidelity of translation. Polyamines canstimulate ribosome subunit association, stabilizethe tRNA structure and reduce the rate of RNAdegradation, enhance both RNA and DNA synthe-sis, help to condense DNA, covalently modifyproteins, and regulate the rigidity and stability ofcellular membranes.10

Certain classes of amines, the catechola-mines, indolamines, and histamine, fulfill im-portant metabolic functions in humans, espe-cially in the nervous system and the control ofblood pressure. Phenylethylamine and tyraminecause a rise in blood pressure, whereas hista-mine reduces it. Histamine has a powerful bio-logical function, serving as a primary mediatorof the immediate symptoms observed in allergicresponses, such as urticarial lesions.121 Althoughbiogenic amines such as histamine, tyramine,and putrescine are needed for many critical func-tions in man and animals, the consumption offood containing high amounts of these may havetoxic effects. Symptoms that can occur after ex-cessive oral intake include nausea, respiratorydistress, hot flush, sweating, heart palpitation,headache, bright red rash, oral burning, andhyper- or hypotension.

The most notorious food-borne intoxicationscaused by biogenic amines are related to hista-mine. Food poisoning involving the consumptionof food containing abnormally high levels of hista-mine has been recognized for many years. Numer-ous outbreaks of histamine poisoning have oc-curred after eating cheese or fish. Histamine is themost toxic amine detected in food (fish, cheese,wine, meat products), and normally it is associated

with food spoilage. The toxic effect depends onhistamine intake, presence of other amines,aminooxidase activity, and the intestinal physiol-ogy of the individual. The gut can absorb six to tentimes more histamine when putrescine, cadaver-ine, and spermidine are present.10

The biogenic monoamines (i.e., serotonin,phenylethylamine, and tyramine) and the di-amines (i.e., histamine and tryptamine) are alsoformed and degraded during normal cellular me-tabolism, playing a variety of physiologicalroles, such as the regulation of body tempera-ture, stomach volume, and pH. They can also af-fect brain activity.10 Biogenic amines such asputrescine, cadaverine, spermidine, and othershave been reported to be radical scavengers.Tyramine has a marked antioxidative activity,which increases with tyramine concentration,and this antioxidative effect may be attributableto its amine and hydroxy groups.141 Polyamines,spermine, spermidine, and putrescine, inhibit theoxidation of polyunsaturated fatty acids, and thisantioxidative effect is correlated with the num-ber of amine groups in the polyamine.70

Nitrosable secondary amines (i.e., agmatine,spermine, spermidine) can form nitrosamines byreaction with nitrite and produce carcinogeniccompounds.37 In general, N-nitroso compoundscan be formed by the interaction of amino com-pounds with nitrosating reagents such as nitriteand nitrogen oxides during the storage, preserva-tion, and cooking of foods.44 The reaction ofnitrosating agents with primary amines producesshort-lived alkylating species that react withother components in the food matrix to generateproducts (mainly alcohols) that are devoid oftoxic activity at the concentrations produced.The nitrosation of secondary amines leads to theformation of stable N-nitroso compounds; thatof tertiary amines produces a range of labile N-nitroso products. Secondary amines such as pu-trescine and cadaverine can react with nitrite toform heterocyclic carcinogenic nitrosamines,nitrosopyrrolidine and nitrosopiperidine.46 Forthis reason, concern over the use of nitrite haspromoted many investigations.43'44'93'108'128

Fortunately, a fairly efficient detoxificationsystem exists in the intestinal tract of animals,

which is capable of metabolizing normal dietaryintakes of biogenic amines. Under normal condi-tions in humans, exogenous amines absorbedfrom food are detoxified rapidly by the action ofamine oxidases or by conjugation; but in aller-genic individuals, or if monoamine oxidase in-hibitors are applied or when intake levels ofamines are too high, the detoxification process isdisturbed and biogenic amines accumulate in thebody. Amino oxidases are induced in the presenceof mono- or diamines.18 The enzymes monoamineoxidase (MAO) and diamine oxidase (DAO) playan important role in this detoxification process.However, when the activity is suppressed by oneor more substances known as potentiators, detoxi-fication is inhibited. This may explain why spoiledfish or aged cheese is more toxic than histamine inaqueous solution.

Alcoholic beverages are good potentiators forsensitivity against biogenic amines. Certaindrugs have been implicated as contributing fac-tors in cases of histamine poisoning. Antihista-mines, antimalarials, and other medications caninhibit histamine-metabolizing enzymes. 18

Some amines, specifically putrescine and cadav-erine, inhibit histamine-detoxifying enzymes.Other amines that may act as potentiators in-clude tyramine (which can inhibit MAO) andtryptamine (which inhibits DAO). Phenylethyl-amine is a DAO and HMT (histamine N-methyl-transferase) inhibitor.121 Some microorganisms(Sarcina lutea, Lactococcus lactis, L. lactissubsp. lactis, and L. lactis subsp. diacetilactis41)contain monoamine oxidase or diamine oxidase(L. lactis subsp. cremoris,41 A. niger, and Tri-chosporon spp112). Some types of yeasts found inRoquefort cheese are capable of assimilating ca-daverine, putrescine, and histamine.13

The levels reported for histamine and its po-tentiators in food would not be expected to poseany problem if normal amounts were consumed.In susceptible individuals, 3 mg of phenylethyl-amine causes migraine headaches; 6 mg totaltyramine intake was reported to be a dangerousdose for patients receiving monoamine oxidaseinhibitors.110 A level of 1000 mg/Kg (amine/food) is considered dangerous for health. Thislevel is calculated on the basis of food-borne his-

tamine intoxications and the amine concentrationin the food. The discrepancy in the toxic histaminelevel in food might be due to the absence or pres-ence of other synergistic biogenic amines like pu-trescine and cadaverine. The European Union hasproposed that the average content of histamineshould preferably be less than 100 ppm, but shouldnot be higher than 200 ppm in fish or fish productsbelonging to the scombridae and clupedae fami-lies.71 The European Union set a three-class planfor maximum allowable levels of histamine infresh fish (n = 9 number of units to be analyzedfrom each lot; c = 2 number of units allowed tocontain a histamine level higher than m = 100ppm; maximum permissible level M = 200 ppm)and fish products (m = 200 and M = 400 ppm).

Conclusions

A reduction in the biogenic amine content offood can be achieved by the following steps.

1 . Select raw material without high levels ofamines or microbial contamination.

2. Provide systematic control at each pro-cessing step with good hygienic practice.This can result in an optimized processwith the corresponding inhibition of spoil-ing microorganisms and low levels of bio-genic amines.

3. In fermented food, the selection criteriafor bacteria used for starter cultures shouldinclude analysis of potential amine pro-duction.

4. Taking into account the effect of potentia-tors (including alcohol and other amines)on the toxicological activity of amines,amine profiles should be determined in-stead of simply the levels of histamine andtyramine.

5. The original concentration of amines infood can change as a result of storage con-ditions and these should be controlled.

ETHYL CARBAMIDES

Ethyl carbamate is an ester of cabamidic acid.Whereas carbamidic acid is unstable and disso-ciates spontaneously into carbon dioxide and

ammonia, the esters of carbamidic acid arestable.

H2N-CO-OH -> NHs + CO2

carbamidic acid ammonia carbon dioxide

The carbamides can be esterified with differentalcohols such as ethanol to form ethyl carbamate.

Formation

The carbamide or carbamidic acid is formedfrom carboxylic and amino groups. The mostcommon carbamide is urea, the final product ofprotein metabolism in ureolytic animals.

HO-CO-OH + NHs ->carbonic acid ammonia

H2N-CO2-NH4 + -* H2N-CO- NH2

ammonium carbamate urea

In fermented beverages, ethyl carbamate isproduced during the fermentation process as a

product of yeast metabolism. Carbamyl phos-phate is produced from adenosine triphosphate(ATP), carbon dioxide, and ammonia by car-bamyl-phosphate synthase, a reaction necessaryfor arginine synthesis, which is enhanced whenammonia levels are high. This compound may re-act with the ethanol that is formed in fermentationand results in ethyl carbamate, which is a potentcarcinogen for humans. Ethyl carbamate can alsobe formed in fermented food (Figure 6-2) by thereaction of ethanol and urea that is produced natu-rally from amino acids like citrulline, which isformed via arginine degradation in wine by someLAB during malo-lactic fermentation;36 or whenurea is added as a yeast nutrient.53

The overall reaction for these compounds is viaan acid-catalyzed ethanolysis. In wines fromgrapes with low nitrogen content and the additionof sugar and water (to further dilute the significantamino acids), the urea remaining in the wine may

arginine (fumaric acid)

(argininosuccinic acid)

(aspartic acid)(citrulline)

(carbamyl phosphate) (ornithine)

carbamyl phosphate + C2H5OH (ethanol) -> C2H5 — COONH2 (ethyl carbamate) + H3 PO4

Figure 6-2 Formation of ethyl carbamate via arginine degradation

be negligible.120 The extent of ethyl carbamate for-mation is likely to depend on the level of precur-sors, the yeast strain used, fermentation condi-tions, and the pH during the process.85

Ethyl carbamate was not detected in a rangeof fermented foods where yeasts were not in-volved as a major population. Experiments on amodel grape juice demonstrated that N-car-bamyl-amino acid was produced by yeast duringfermentation. Because the conversion of N-car-bamyl-amino acid into ethyl carbamate is not fa-vored chemically, N-carbamyl amino acids arenot expected to contribute substantially to theformation of ethyl carbamate in wines undernormal conditions of wine making and aging.45'90

Ingledew et at.53 demonstrated that yeast fer-mentation itself did not lead to the production ofethyl carbamate. However, when subsequentheating steps were applied to fermented bever-ages in which urea was present or made, ethylcarbamate was formed in significant quantities.

The source of urea is arginine breakdown viaarginase activity. Citrulline and other N-car-bamyl amino acids can react with ethanol toform ethyl carbamate. Samples of commerciallyprocessed juices were fermented under con-trolled conditions to determine variation inamino acid uptake or excretion during yeastgrowth and fermentation. There was almost al-ways rapid uptake of all of the amino acids ex-cept alanine, proline, citrulline, ornithine, andarginine. Urea excretion and uptake were depen-dent primarily on yeast strain and on amounts ofarginine remaining in the juice during fermenta-tion. Amounts of urea excreted tend to increaseat higher temperatures. At very high amino acidlevels, arginine is metabolized very slowly,

which minimizes urea excretion. Comparison ofvarious yeasts shows that patterns of urea uptakeare generally similar.90

Studies on factors (Table 6-3) influencing theformation of ethyl carbamate in wines havedemonstrated that ethyl carbamate can beformed spontaneously from urea, citrulline, orcarbamyl phosphate in acidic ethanolic solu-tions. At high temperatures, all compounds con-taining a carbamyl structure have the potential toform ethyl carbamate. Although all wines havethe potential to form ethyl carbamate, the likeli-hood is greatest in wines that have undergonemalo-lactic fermentation.36 Nitrogenous con-stituents of wine other than urea are not corre-lated with urea formation, and N-fertilization ofthe grapes does not affect urea formation in theresulting wines.119 In Cabernet Sauvignon andChardonnay wines, ethyl carbamate contents in-creased over two years of storage. There was nosignificant effect of wine type (red vs. white) orpH on ethyl carbamate content. Holding tem-perature had a significant effect on ethyl car-bamate formation; increasing the temperature by10 0C may increase the rate of ethyl carbamateformation up to 3 -fold. Thus, knowledge of theurea concentration and wine storage temperaturepermits a good estimation of the amount of ethylcarbamate that will form over a period represen-tative of wine storage.120

Applications

In the past, the American baking industry hasused several reducing agents such as ascorbicacid, sodium metabisulphite, and cysteine-N-carbamide as chemical dough developers in-

Table 6-3 Factors Affecting Ethyl Carbamate (EC) Synthesis

Factors Causative Effect References

Substrate Free urea, carbamyl-P (increase EC) 36, 61,137Enzyme activity Carbamyl-P synthetase, urease (increase EC) 61pH Acid (increases EC) 61Heat process Toasting (increases EC) 21Ascorbic acid Decreases EC 21

stead of cysteine. Cysteine-N-carbamide hasbeen shown to improve sour bread dough notice-ably, and the addition of cysteine to the sour-dough system significantly reduced mixing re-quirements and increased bread volume.130 Inchemical dough development, cysteine is usedmost commonly. It speeds the breadmaking pro-cess by expediting disaggregation by splittingdisulphide bonds between protein aggregatesthrough reduction and promoting disulphide inter-change with less mixing. Carbamide considerablyincreases the stability of starch phosphate at lowtemperatures. An addition of a small quantity ofcarbamide to the starch increases its water absorp-tion power, raises its viscosity, and improves thetransparency of its aqueous solution.

In the European Union, the use of cysteine-N-carbamide is forbidden because of its possiblerole in promoting carcinogenesis. For this rea-son, when it is used in any food-related material,it must be within the limits permissible by thesanitary authorities. The use of carbamide as asubstrate in the growth of yeasts led to improvedmolasses fermentation. Carbamide addition as anitrogen source was recommended for molasseswith reduced nitrogen content. Molasses-grownyeast reduced leavening time in the baking pro-cess and improved the quantity of molasses alco-hol with simultaneous production of baker'syeasts of good quality.102 After molasses fer-mentation is finished, carbamide can be re-moved from alcoholic beverages with urease,preferably from LAB; the optimum for this reac-tion is pH 2-5 .55 Carbamide was the most effi-cient nitrogenous substance added to manioc(cassava) mash for the nutrition of the yeast cellsin the saccharification and fermentation of man-ioc starch.84

Carbamides also find application in packag-ing material for maintaining more moisture inthe product.6'1 17In other cases, carbamide isadded in the form of carbamide peroxide to givehydrogen peroxide in processes for protectingmilk or milk products containing lactoperoxi-dase from bacterial spoilage during storage.15

Carbamide used as a feed supplement or as fer-tilizer can influence the chemical composition ofvegetables and meat.136 When carbamides are

used as insecticides in wheat storage, the controland norms to be followed in the fumigation andstorage of wheat to be used as seed or for foodare very strict. At certain levels, the presence ofcarbamide in some foods can be detected by itsodor or bitter taste.5

Health Effects

Ethyl carbamate (urethane) has mutagenicand carcinogenic properties. Due to the toxicityof carbamide, there are maximum permissibleconcentrations in some countries set on the basisof animal trials in foods and water. The presenceof various levels of ethyl carbamate in distilledspirits and wines, and fermented foods such asbread, soy sauce, miso, and yeast spread hasbeen reported. In most countries, there is no le-gal limit for ethyl carbamate, but the Food andAgricultural Organization (FAO)AVorld HealthOrganization (WHO) suggested a level of 10ppb for soft drinks; in Canada, the tolerancelevel ranges from 30 ppb in wines to 400 ppb indistilled spirits.85

Conclusions

Ethyl carbamate is a carcinogenic and mu-tagenic compound that results from the esterifi-cation of carbamic acid by ethanol in alcoholicfermentations. The use of proper starter culturesof yeasts in the manufacturing of fermentedfoods can result in increased levels of safety ofthese foods. The selection of nontoxigenic start-ers that antagonize pathogenic microorganismsand have detoxifying ability should be a priority.

NITROSAMINES

Nitrosamines are formed by the reaction ofsecondary and tertiary amines or amino groupscontained in compounds such as dialkylamines,alkylarylamines, piperazine, and pyrrolidine,with nitrite. Two precursors, secondary aminesand nitroso acid, are required for the formationof nitrosamines. Certain foods have been foundto contain various nitrosamines, many of whichare known carcinogens. The formation of nitro-

samine has been demonstrated in food-nitritemixtures. Although amines are present in foods,wine, tobacco products, drugs, and other envi-ronmental chemicals such as pesticides, sourcesof nitrite, the other precursor, vary. Human ex-posure can be by the ingestion or inhalation ofpreformed N-nitroso compounds or by endog-enous nitrosation. It has been established un-equivocally that N-nitroso compounds are alsoformed in the body from precursors that arepresent in the normal diet.44

Formation

The formation of nitrosamine from nitritetakes place through various steps, as indicated inFigure 6-3.

The rate constant K is independent of pH, butthe amine and nitrous acid concentrationschange with pH. The first two reactions are fa-vored by acidic conditions. However, if the envi-ronment is too acidic, the amine will be proto-nated and unable to react with the nitrousanhydride (N2Cb). Nevertheless, nitrosation canalso occur under the nonacidic conditions thatare found in some foods, and oxides of nitrogen(NOx) can react directly without the requirementfor acid conditions. Some components found infoods catalyze nitrosation reactions. Nitrite infood can be produced by microbial nitrate reduc-tion. Nitrates are naturally present in soil, water,and food, and also in the air at low concentrationas a result of industrial pollution. Nitrates in soil

can accumulate due to fertilizer treatments. Forexample, under conditions of excessive fertiliza-tion at low temperatures and with low intensityof light, nitrate can accumulate in vegetables.33

Nitrosamines can become a part of foods bythe use of nitrite as an additive, through process-ing, especially drying, and through their forma-tion during food preparation, migration fromfood contact surfaces, or indirect addition alongwith other additives. Products made from rubber(elastomers) contain volatile nitrosamines thatcan migrate to food or drinks. The rubber nettingused to hold hams and other cured products dur-ing processing can also result in the migration ofvery small amounts of nitrosamines to foods.Fiddler et a/.30 confirmed these findings and re-ported that additional nitrosamines could beformed when rubber products were exposed tosolutions that were similar to human saliva.There is a controversy regarding the possiblemigration of nitrosamines from rubber elasticnetting into meat products. Rubber elasticizednetting has been in use since the 1960s, and theKeystone Casing Supply Company (Carnegie,Pa) first received confirmation of approval fortheir Jet Net formulation netting from the U.S.Department of Agriculture (USDA) in 1964. In1985, a study showed possible migration of nit-rosamines from the nipples on baby bottles intosaliva or milk.105 To minimize the potential fornitrosamines formation, the manufacturers ofelastic rubber netting reformulated the rubbercomponent to reduce the level of nitrosatable

NITROUSNITRITE NITROUSACID ANHYDRIDE WATER

fast

NITROUS SECONDARYANHYDRIDE AMINE NITROSAMINE NITROUSACID

Figure 6-3 Formation of nitrosamine from nitrite at a rate of K (R2NH) x (HNO2).

slow

amines.75 Some paper-based packaging materi-als contain trace levels of volatile nitrosamines,which can migrate to foods.44

Nitroso compounds can be formed inside thebody, and the diet influences the nature of nitrosocompound formed. Humans are also exposed tonitrosamines through normal physiological pro-cesses in the stomach through the reaction be-tween amines and nitrite. The nitrite can be de-rived from two sources: consumption of foods thatcontain nitrite and the conversion of nitrate to ni-trite in the body. Nitrate can be derived both fromnormal nitrogen metabolism and from the diet.Dietary nitrate is absorbed quickly after ingestionand is circulated in the blood. Then it is excretedand concentrated in the saliva, where microorgan-isms in the mouth convert it to nitrite. The nitrite isswallowed, resulting in the formation of nitro-samines in the stomach. The major source of di-etary nitrate is vegetables, with cured meats con-tributing a small amount. Diet may play a role inthe types and amounts of nitrous compounds thatare formed endogenously. The dietary balance be-tween nitrate and ascorbic acid may be particu-larly important because ascorbic acid is capable ofinhibiting endogenous nitrosamine formation re-sulting from nitrate exposure.44

Nitrosamines in Food

Cured Food

The origin of the use of nitrates as a meat cur-ing agent is lost in antiquity. Humans have beenusing salt for preserving food for thousands ofyears, obtaining good products with flavor andcolor typical of cured foods. Common salt usu-ally contains impurities from other salts, includ-ing nitrates, which give the cured color and thecured aroma. The term "curing" came to be usedfor the first time in northern Germany to de-scribe the treatment of meat and fish with com-mon salt. This has meant that even today, there issome confusion between salting, treating withcommon salt, and curing, treating with nitrate ornitrite-curing salts. The salt used to preservemeat contains potassium nitrate, which producesthe characteristic flavor and color of the curedmeats. Later, scientific understanding of the cur-

ing process began to evolve, and the role of ni-trite was recognized. It was observed that nitratewas reduced to nitrite by bacterial activity, andthat nitrite is reduced to nitric oxide to form thetypical curing color. These findings led to theuse of nitrite directly as a curing ingredient. Ni-trite is a highly reactive chemical. In the com-plex meat system, several reactions can occur,including the reaction of nitric oxide with myo-globin to form the red color of the cured meat.

Concern over nitrosamines stimulated consid-erable interest in the value of nitrite as an antimi-crobial agent. This has focused mainly on theminimum levels necessary for the control of C.botulinum and on developing alternatives to re-place nitrite wholly or in part. Hauschild38 sug-gested that, from a safety point of view, signifi-cant reductions in nitrite input could be made fora number of cured meats without the need forcompensation. Fermented meats may belong tothis group. With respect to the antimicrobial ac-tivity of nitrite on other toxigenic or infectiousmicroorganisms in food, the evidence is rathercontradictory due in part to the complex interac-tions occurring during product processing. Dur-ing the 1970s, as pressure mounted to minimizethe allowable level of nitrite for curing meat,there was concern about the risk of botulism.This concern resulted in a large number of col-laborative studies conducted jointly by govern-ment and industry. The general conclusion wasthat nitrite is required to maintain safety, but itwas also noted that the interaction of other fac-tors such as salt level, pH, and heat treatment iscritical. In 1982, the Committee on Nitrite andAlternative Curing Agents in Food82 issued a re-port in which the status of the search for alterna-tives to nitrite was reviewed exhaustively. Theylisted the following potential alternatives and/orpartial substitutes: ascorbate and tocopherol, ir-radiation, LAB, potassium sorbate, sodiumhypophosphite, and fumarate esters. The amountof nitrite added to the bacon can be reduced bythe addition of LAB and fermentable sugar,which lower the pH, thereby ensuring that C.botulinum will not grow and produce toxin.38'124

N-nitrosation is influenced by many factorssuch as pH, level of precursors, basicity of the

nitrosatable amines, and the presence of cata-lysts and inhibitors. A number of compoundssuch as sulfur dioxide, bisulfite, oc-tocopherol,ascorbic acid, and glutathione are known to in-hibit N-nitrosation. Simple phenols andpolyphenolic compounds can decrease or in-crease the rate of N-nitrosation reactions de-pending on their structure and reaction condi-tions, especially pH. Under acidic conditions,certain phenolic compounds are known to com-pete more efficiently for the available nitritethan the amines. The inhibition of nitrosationoccurs in most cases because the nitrosatingagents are converted to innocuous products suchas nitric oxide, nitrous oxide, or nitrogen.111

Vegetable juices inhibit N-nitrosodimethyl-amine formation to a certain degree, but lemonjuice was found to inhibit its formation to thegreatest extent. The inhibition of N-nitro-sodimethylamine formation by a solution con-taining the same amount of ascorbic acid as thatof the lemon juice was 0.6%, suggesting that an-other compound is responsible for the inhibi-tion.4 Ascorbic acid (vitamin C), tocopherol (vi-tamin E), and erythorbic acid (isoascorbic acid)are used routinely in meat curing, primarily tospeed up the curing process.91'98'107 Ascorbic anderythorbic acids have the disadvantage of beingwater soluble, whereas nitrosation occurs prima-rily in the lipid phase.44 Tocopherols have thedisadvantage that they are lipid soluble and,hence, difficult to disperse in an aqueous curepump mix. Tocopherol can be coated onto thesalt, when salt dissolves in the brine, a fine sus-pension of tocopherol results.

Temperature influences nitrosamine forma-tion in food that is rich in amines and nitrites.34

The methods used for heating also have an influ-ence. Cooking by broiling and frying, for ex-ample, generates a higher level of nitrosaminethan boiling and microwave heating.89 Cookingcan also generate additional amine precursors tocontribute to preformed nitrosamines in foods.31

Smoked products have a higher level of nitro-samines due to the high temperatures and thecomponents of smoke.96'97 Interest has been fo-cused on cured meats because the potential fornitrosamine formation would be greatest in

foods to which nitrite had been added, and be-cause the ingestion of nitrite might lead to theformation of nitrosamines in the stomach. Nitro-samines have been found consistently in bacononly after it has been fried. Fried bacon containstwo or three volatile nitrosamines, N-nitro-sopyrrolidine and N-nitrosodymethylamine, N-nitrosothiazolidine at lower levels, as well asnonvolatile nitrosamine (e.g., N-nitrothiazoli-dine carboxylic acid) in lower concentration andless frequently. When bacon is fried, nitrosatingagents cannot be extracted from the raw or friedout fat, but the nitrogen oxides formed from theadded nitrite react with the unsaturated lipidsduring the curing process to form compoundsthat can decompose during frying to formnitrosating agents (e.g., NO, NCh, N2Cb).

N-nitrosamine formation occurs primarily inthe lipid phase of bacon and can be effectivelyinhibited by lipophilic free-radical scavengers.This implies a free-radical mechanism that maynot be a direct transfer of a nitroso group be-tween the nitrosating agent and the amine.14 Thenitro-nitroso compounds demonstrated thegreatest capacity for N-nitrosation. Nitrogen di-oxide can also react with unsaturated fatty acidsto form uncharacterized nitrosating agents.78

Dinitrogen trioxide (№03) easily reacts withmethyloleate to form several additional prod-ucts, some of which are capable of N-nitrosationunder conditions similar to frying bacon.101 Sev-eral factors influence nitrosamine formation infried bacon, including the age and fat content ofthe bacon when cooked, the amount of nitriteused in the cure, the presence of nitrosation in-hibitors, and the temperature of frying. The olderthe raw belly bacon, the greater protein break-down is to secondary amines. Bacon with higherfat content tends to produce more nitrosamineswhen fried. In general, bacon that is fried atlower temperatures contains less nitrosamines,even if it is cooked well done. When bacon isprepared in a microwave oven, it contains lessnitrosamine than when it is fried because of thelower temperature. Nitrosamines have not beenfound routinely in other meats, even when fried.This is because most cured meats, with the ex-ception of bacon, are not fried to complete dry-

ness. The volatile nitrosamines formed in othermeat products are carried off with the water va-por during the frying process.44

Other factors such as age of food, storage, andprocessing may affect the amount of amine pre-cursors available. In the Turkish fermented sau-sage, sucuk, nitrite concentration decreased inall raw samples during storage for up to 10 days.Nitrosamines were not detected in either raw orcooked sucuk made with 150 ppm nitrite. Rawsamples made with 200 ppm nitrite contained0.015 ppm N-nitrosodiethylamine and 0.057ppm N-nitrosodimethylamine; however, no nit-rosamines were detected in cooked samples ofsucuk made with 200 ppm nitrite. Sucuk madewith 300 ppm nitrite was free from nitrosaminesbefore cooking, fried samples contained up to0.034 ppm N-nitrosodiethylamine, and grilledsamples contained up to 0.011 ppm N-nitro-sodiethylamine.24

Amines have been historically associatedwith fish and fish products with respect to dete-rioration of quality. These amines increase thelikelihood of nitrosamine formation in the pres-ence of nitrite in curing fish. With fresh fish,cooking increased or caused N-nitrodimethyl-amine to be formed where none was presentoriginally. This effect was more pronouncedwith salt-dried seafoods, especially when theywere broiled with gas. The concentration ofamines present, particularly dimethylamine, apotential precursor of N-nitroso dimethylamine,varies with the fish species and other factors.31

Other fermented foods in which nitrosamineshave been detected are alcoholic beverages suchas brandy, vodka, red wine, and vermouth,3 aswell as Bulgarian hard and semihard cheeses.62

Noncured Products: Beer

There are many studies on the presence of ni-trosamines in beer or in hops, malt, and brewingmaterial.54'56'116'129'131 The presence of nitrogenoxides (NOx) in the air in malt kilns may result inthe formation of dimethylnitros amine in themalt, and carryover of nitrosamine into thebeer.34 During the malt drying by the direct-firekilning process, N-nitrosodimethylamine is pro-duced. Oxides of nitrogen formed from combus-

tion become incorporated into the drying air,where they react with amines in the malt. In re-sponse to these findings, the malting industrysignificantly reduced N-nitrosodimethylamineformation in the process by converting to indi-rect-fire kilns and/or by introducing sulphur di-oxide during kilning.106 Hordenine and gramine,tertiary amine alkaloids that are biosynthesizedfrom tyrosine in malt, have been demonstratedas amine precursors for N-nitrosodimethyla-mine95 (Figure 6-4).

Nitrosamines can also rise to dangerous levelsin drinking water,58'72 snuff of tobacco,88 andnicotine chewing gum.88

Health Effects

Quite large quantities of nitrate are ingesteddaily with food, particularly from vegetablesand, locally, from drinking water. Epidemiologi-cal studies have demonstrated a correlation be-tween the concentration of nitrate in potable wa-ter and the death rate due to stomach cancer.33

There is very little nitrate in meat; the same isgenerally true of meat products. In an overallbalance sheet for human nitrate ingestion, meatand meat products play only a subordinate role.As far as nitrite is concerned, however, it isknown to have a harmful effect on the humanorganism. By attaching itself to hemoglobin, theblood pigment, nitrite blocks oxygen supplies inthe body. In the 1920s, nitrite was still freelyused, added in the wrong dose or inadvertentlyconfused with common salt. As a result of thiserroneous application, people died by ingestingnitrite from cured meat products. Because ofthis, the nitrite regulations were issued at the be-ginning of the 1930s, banning the use of pure ni-trite in cured meat products. Since then, its usehas only been permitted in a mixture with com-mon salt. The curing of meat products became asubject of further concern in the 1970s, however,when it became known that nitrosamines aresometimes produced in meat products and canbe harmful to human health.

Most nitrosamines are very carcinogenic. Thequantity of nitrosamine produced in meat prod-ucts is variable. But even doses of carcinogenic

or toxic substances that lie below the thresholdmay contribute to ill health by their cumulativeeffects. This should be remembered when curingsalt is used. Even residues involving only aslight risk from the toxicological point of viewshould be regarded as undesirable and avoidablein meat and meat products. It has only been in ex-ceptional cases that tests have been made to deter-mine to what extent a foreign substance is reab-sorbed in the human gastrointestinal tract afterbeing bound to muscle protein or animal fat. Sub-stances that have a close binding affinity to mac-romolecular structures (e.g., ribonucleic acids,proteins) are likely to be reabsorbed only incom-pletely after ingestion with a full mixed diet.This means that low "secondary" bioavailabilityis a further safety factor for the consumer.32

Epidemiological studies do not produce hardexperimental data that prove specific cancer-causative factors. Rather, they provide associa-tion hypotheses or "risk factors" for furtherstudy. It is well known that nitrosamines are abroad class of compounds that are formed fromthe nitrosation of substituted amides, urea, car-bamates, and guanidine.9'43 The nitrosamides aredirect active carcinogens, meaning that the acti-

vation is nonenzymatic, occurring by spontane-ous hydrolysis. Bioactivation of nitrosamines,on the other hand, occurs through an initial 2-hydroxylation, catalyzed by cytochrome P450(Figure 6-5). This distinction between these twoclasses of N-nitroso compounds correlates with atendency for nitrosamides to produce tumors at theinitial organ exposed (e.g., stomach), whereas nit-rosamines can initiate tumors at distal sites. Ap-proximately 300 different N-nitroso compoundshave been evaluated for carcinogenicity, and morethan 90% have been found positive.8

Human exposure to nitroso compounds oc-curs through three main routes.

1. Exogenous levels in foods. These are usu-ally nitrosamines, as the nitrosamides tendto be unstable. The formation of nitro-samines in foods occurs through the reac-tion of secondary or tertiary amines with anitrosating agent (commonly, nitrous an-hydride), which in turn is formed from so-dium nitrite added to a number of foods asa preservative and color enhancer.

2. Tobacco smoke. Tobacco smoke containstobacco-specific nitrosamines, which

N - METHYLTYRAMINE HORDENINE

N - METHYL - 3 - AMINOMETHYLINDOLE

Figure 6-4 Malt alkaloids.

GRAMINE

DIMETHYLAMINE + NITRITE DIMETHYLNITROSAMINE

CH3- Guanlne

O - 6 - METHYLGUANINE

Figure 6-5 Metabolism of food-borne procarcinogens. Note: a = cytochrome P450 2El; s = spontaneousreaction.

have been implicated in the etiology ofcancers associated with smoking.39 TheU.S. National Academy of Sciences81 esti-mated that exposure of smokers to volatilenitrosamines through cigarette smoke (17|Lig/day) is two orders of magnitude higherthan the estimated dietary exposure due toingestion of food with the highest nitro-samine level (fried bacon, 0.17 jig/day). Infact, foods as sources of human exposureto nitrosamines were ranked lower than ei-ther automobile interiors (0.5 |ig/day) orcosmetics (0.41 |Lig/day).

3. Endogenous formation in the acidic envi-ronment of the stomach. Vegetables con-tain high levels of nitrate, whereas dietarynitrite comes primarily from cured meats.However, the dietary contribution of ni-trite is small compared to the conversionof nitrate to nitrite by microorganisms ofthe gastrointestinal tract.

In order to control the possible health risks ofpreservatives, it is of primary importance thatthe consumption of preservatives in foods mustbe in accordance with the acceptable daily intakebased on the highest intake that does not giverise to detectable adverse effects. The values areoften derived from studies with laboratory ani-mals and a safety factor is then applied to thehighest level at which no adverse effect is ob-served. Exposure to food preservatives is un-

likely to be constant and will change in the fu-ture as a consequence of technological develop-ments in the food industry and changes in di-etary habits. There will therefore be a continuingneed to update exposure data to monitor the ef-fects of these developments on preservative in-takes. The formation of carcinogenic N-nitro-samines from the administration of nitrite andsecondary amines in animal studies only occursat nitrite levels that are far in excess of the di-etary intakes that humans are exposed to. In theUnited Kingdom, the Scientific Committee forFood has recommended that nitrate should notbe used as an additive in infant foods in view ofthe acute effect of infantile methemoglobinemia.Although the nitrate anion is chemically verystable, it is readily reduced microbially to nitrite;approximately 5% of ingested nitrate is so re-duced to nitrite. In practice, however, epidemio-logical investigations have failed to demonstratean unequivocal link between nitrate exposureand cancer.75

Conclusions

Nitrite is one of the most important food ad-ditives from both an economic as well as a tech-nical standpoint. The addition of nitrite changesperishable meats and fish into unique curedproducts such as bacon and ham, which havedesirable flavor and appearance characteristicsand longer shelf life. It also protects these prod-

ucts against the deadly food-borne bacterium,C. botulinum.

The correct use of nitrite in bacon results inthe occurrence of very low levels of nitro-samines, which at higher levels have beenshown to be carcinogenic in laboratory animals.

REFERENCES

1 . Ababouch, L., Afilal, M. E., Benabdeljelil, H. & Busta,F. (1991). Quantitative changes in bacteria, amino ac-ids and biogenic amines in sardine (Sardinapilchardus) stored at ambient temperature (25-28 de-gree) and in ice. IntJFoodSci Technol 26, 297-306.

2. Abd-Alla, E. A. M., El-Shafei, K., Ibrahim, G. A. &Sharaf, O. M. (1996). Changes in microflora and bio-genic amines of some market processed cheeses duringstorage. Egypt J Dairy Sd 24, 217-226.

3. Acet, H. A., Tras, B., Karahan, I. & Bas, A. L. (1993).Determination of nitrosamines in alcoholic beverages.Doga Turkish J Vet An Sd 17, 275-279.

4. Achiwa, Y., Hibasami, H., Kada, T. & Komiya, T.(1997). Inhibitory effect of vegetable and fruit juiceson formation of N-nitrosodimethylamine. J Jpn SocFood Sd Technol (Nippon-Shokuhin-Kagaku-Kogaku-Kaishi) 44, 50-54.

5. Akachkov,V. P. (1973). Storage of smoked productsfrom cartilaginous fish. Rybnoe Khozyaistvo 3, 75-76.

6. Anisinov, I. G., Vyazum, T. M., Soletskii, O. L,Kostin, N. L, Zakharnchenko, A. D., Ostrikova, L. V.,Shiler, G. G., Rozdov, I. A., Bankaev, E. S., Usmanov,R. M. & Utkin, K. I. (1990). Coating composition forcheese. URRS Patent SU 1.613.097.

7. Askar, A. & Treptow, H. (1986). Biogene amine inLebensmitteln: Vorkommen, Bedeutung undBestimmung. Stuttgart, Germany: Eugen UlmerGmgH and Company.

8. Bailey, G. S., Scanlan, R. A., Selivonchick, D. P. &Williams, D. E. (1991). Food toxicology. In Encyclo-pedia of Human Biology. Vol. 3, pp. 671-681. Editedby R. Dulbelcco. New York: Academic Press.

9. Bailey, G. S. & Williams, D. E. (1993). Potentialmechanisms for food-related carcinogens andanticarcinogens. Food Technol 47, 105-1 18.

10. Bardocz, S. (1995). Polyamines in food and their con-sequences for food quality and human health. TrendFood Sd Technol 6, 341-346.

11. Bauer, F., Seus, I., Paulen, P. & Valis, S. (1994). Theformation of biogenic amines in meat products. InProc. 40th Int Cong. Meat Sd Technol, Vol. 25, pp.l-3. Clermont-Ferrand, France.

Beverages such as beer also contain volatilenitosamines. Through modifications in the pro-cessing and use of chemical inhibitors, and theaddition of lactic starters to meat products, thefood industry has substantially reduced the con-tent of volatile nitrosamines in human food.

12. Bauza, T., Blaise, A., Teissedre, P. L., Mestres, J. P.,Daumas, F. & Cabanis, J. C. (1995). Changes in bio-genic amines content in musts and wines during thewinemaking process. Sciences de Aliments 15, 559-570.

13. Besanc.on, X., Smet, C., Chabalier, C., Rinemale, M.,Reverbel, J. P., Ratomehenina, R. & Galzy, P. (1992).Study of surface yeast flora of Roquefort cheese. Int JFood Microbiol 11, 9-18.

14. Bharucha, K. R., Cross, C. K. & Rubin, L. J. (1985).Ethoxyquin, dihydroethoxyquin, and analogues asantinitrosamine agents for bacon. J Agric Food Chem33, 834-839.

15. Bjorck, K. &Claesson, C. (1980). Method for protect-ing milk from bacterial spoilage by adding thiocyanateand the hydrogen peroxide in the form of alkalipercarbonate, alkali-peroxide or urea hydrogen perox-ide. Swedish Patent 412.838.

16. Boskou, G. & Debevere, J. (1998). In vitro study ofTMAO reduction by Shewanella putrefaciens isolatedfrom cod fillets packaged in modified atmosphere. Int JFood Microbiol 39, 1-10.

17. Bover-Cid, S., Izquierdo-Pulido, M. & Vidal-Carou,M. C. (1999). Effect of proteolytic starter cultures ofStaphylococcus spp. on biogenic amine formation dur-ing the ripening of dry fermented sausages. Int J FoodMicrobiol 46, 95-104.

18. Brink, B. Ten, Damink, C., Joosten, H. M. L. J. & Huisin't Veld, J .H. J. (1990). Occurrence and formation ofbiologically active amines in foods. Int J FoodMicrobiol 1 1,73-84.

19. Buncic, S., Paunovic, L., Teodorovic, V., Radisic, D.,Vojinovic, G., Smiljanic, D. & Baltic, M. (1993). Ef-fects of gluconodeltalactone and Lactobacillusplantarum on the production of histamine andtyramine in fermented sausages. Int J Food Microbiol17, 303-309.

20. Buteau, C., Duitschaever, C. L. & Ashton, G. C.(1984). A study of the biogenesis of amines in a Villardnoir wine. Am JEnol Vitic 35, 228-236.

21. Canas, B. J., Diachenko, G. W. & Nyman, P. J. (1997).Ethyl carbamate levels resulting from azodicarbon-amide use in bread. Food Add and Contam 14, 89-94.

22. Cantoni, C., Bersani, C., Damenis, C. & Comi, G.(1994). Biogenic amines in typical Italian dry sau-sages. /m/ Aliment 33, 1239-1243.

23. Celano, G. V., Cafarchia, C. & Tiecco, G. (1996). For-mation of biogenic amines in Pecorino Pugliesecheese. Ind Aliment 35, 249-252.

24. Develi, N., Gokturk, S. & Ercan, I. (1997). Effects ofcooking methods on N-nitrosamines in sucuk samplescontaining different nitrite concentrations. Il-Kontrol-Laboratuvar-Mueduerluegue-Ankara 3 6 .

25. Diaz-Cinco, M. E., Fraijo, O., Grajeda, P., Lozano-Taylor, J. & Gonzales-de-Mejia, E. (1992). Microbialand chemical analysis of Chihuahua cheese and rela-tionship to histamine and tyramine. J Food Sd 57,355-356, 365.

26. Donhauser, S., Wagner, D. & Geiger, E. (1992). Bio-genic amines: importance, occurrence and evaluation.Brauwelt 132, 1272-1274, 1276, 1278, 1280.

27. Dumont, E., de Geeter, H. & Huyghebaert, A. (1992).Presence and formation of biogenic amines in localBelgian beers. Meded Fac Landbouwwet RijksunivGent 57, 1911-1913.

28. Eerola, S., Maijala, R., Roig-Sagues, A. X., Salminen,M. & Hirvi, T. (1996). Biogenic amines in dry sau-sages as affected by starter culture and contaminantamine-positive Lactobadllus. J Food Sd 61, 1243-1246.

29. Eerola, S., Roig-Sagues, A. X., Lilleberg, L. & Aalto,H. (1997). Biogenic amines in dry sausages duringshelf-life storage. Food Res Technol 205, 351-355.

30. Fiddler, W., Pensabene, J. W., Gates, R. A., Custer, C.,Yoffe, A. & Phillpo, T. (1997). N-nitrosodibenzyl-amine in boneless hams processed in elastic rubbernettings. JAOACInt 80, 353-358.

3 1 . Fiddler, W., Pensabene, J. W., Gates, R. A., Hale, M. &Jahncke, M. (1992). N-nitrosodimethylamine forma-tion in cooked frankfurters containing Alaska pollock(Theragra chalcogramma) mince and surimi. J FoodSd 57, 569-571,595.

32. Fink-Gremmels, J. & Leistner, L. (1986). Comparativetoxicological evaluation of possible residues in meat.Research and practice in food hygiene. In 27th Meetingof the Food Hygiene Section. Garmisch-Parten-kirchen, 9-12 Sept. 1986.

33. Garcia-Roche, M. O. & Hernandez, A. M. (1989).Sources and presence of, and risk from nitrates and ni-trites: situation in Cuba. Alimentaria 199, 67-71.

34. Gifhorn, A. & Meyer-Pittroff, R (1997). Measurementof NOx contamination as a measure for prevention ofnitrosodimethylamine formation in kilning of malt.Brauweltlll, 1149-1153.

35. Giraffa, G., Carminati, D. & Neviani, E. (1997). En-terococci isolated from dairy products: a review ofrisks and potential technological use. J Food Prot 60,732-738.

36. Granchi, L., Paperi, R., Rosellini, D. & Vincencini, M.(1998). Strain variation of arginine catabolism amongmalolactic Oenococcus oeni strain of wine origin. Ital-ian J Food Sd 10, 351-357.

37. Halasz, A., Barath, A., Simon-Sarkadi, L. & Holzapfel,W. (1994). Biogenic amines and their production by mi-croorganisms in food. Trend Food Sd Technol 5, 42-49.

38. Hauschild, A. H. W. (1982). Assessment of botulismhazards from cured meat products. Food Technol 36,95-97.

39. Hecht, S. S. & Offtnan, D. (1988). Tobacco-specificnitrosamines: an important group of cancerigens in to-bacco and tobacco smoke. Cardnigenesis 9, 875-884.

40. Hernandez-Herrero, M. M., Roig-Sagues, A. X.,Rodriguez- Jerez, J. J. & Mora- Ventura, M. T. (1999).Halotolerant and halophilic histamine-foraiing bacte-ria isolated during the ripening of salted anchovies(Engraulis encrasicholus). J Food Prot 62, 509—514.

41. Holt, J. G., Krieg, N. R., Sfheath, T. H. A., Stanley, J.T. & Williams, S. T., eds. (1994). Sergey's Manual ofDeterminative Bacteriology, 9th edn. Baltimore: MD:Williams & Wilkins.

42. Hornero-Mendez, D. & Garrido-Fernandez, A. (1997).Rapid high performance liquid chromatography analy-sis of biogenic amines in fermented vegetable brines. /Food Prot 60, 414-419.

43. Hotchkiss, J. H. (1989). Preformed N-nitroso com-pounds in foods and beverages. A dv Food Res 3 1 , 53—115.

44. Hotchkiss, J. H. & Cassens, R. G. (1987). Nitrate, ni-trite, and nitroso compounds in foods. Food Technol41, 127-134.

45. Huang , Z. & Ough, C. (1993). Identification of N-Car-bamyl amino acids in wine and yeast cells. Am J EnolVitic44, 1,49-55.

46. Huis in't Veld, J. H. J., Hose, H., Schaafsma, G. J.,Silla, H. & Smith, J. E. (1990). Health aspects of foodbiotechnology. In Processing and Quality of Foods.Vol. 2, Food Biotechnology: Avenues to Healthy andNutritious Products, pp. 2.73-2.97. Edited by P.Zeuthen, J. C. Cheftel, C. Ericksson, T. R. Gormley, P.Link & K. Paulus. London: Elsevier Applied Science.

47. Hwang, D. F., Chang, S. H., Shian, C. Y. & Cheng, C.C. (1995). Biogenic amines in the flesh of sailfish(Istiophorus platypterus) responsible for scombroidpoisoning. J Food Sd 60, 926-928.

48. Ibe, A., Kamimura, H., Tabata, S., Hayano, K.,Kimura, Y. & Tomomatsu, Y. (1996). Contents of non-volatile amines in fermented foods. II. TAnnu Rep To-kyo Metrop Res Lab Publ Health 47, 90-94.

49. Ibe, A., Nishima, T. & Kasai, N. (1992). Bacteriologi-cal properties of and amine-production conditions fortyramine and histamine-producing bacterial strainsisolated from soybean paste (miso) starting materials.Jpn J Toxicol Environ Health 38, 403-409.

50. Ibe, A., Nishima, T. & Kasai, N. (1992). Formation oftyramine and histamine during soybean paste (miso) fer-mentation. Jpn J Toxicol Environ Health 38, 181-187.

51. Ibe, A., Saito, K., Nakazato, M., Kikuchi, Y.,Fujinuma, K. & Nishima, T. (1991). Quantitative de-termination of amines in wine by liquid chromatogra-phy. J Assoc Off Anal Chem 74, 695-698.

52. Ibe, A., Tamura, Y., Kamimura, H., Tabato, S.,Hashimoto, H., Iida, M. & Nishida, T. (1991). Deter-mination and contents of non-volatile amines in soy-bean paste and soy sauce. Jpn J Toxicol EnvironHealth 37, 379-386.

53. Ingledew, W. M., Magnus, C. A. & Patterson, J. R.(1987). Yeast foods and ethyl carbamate formation inwine. Am JEnol Vitic 38, 332-335.

54. Izquierdo-Pulido, M., Hernandez- Jo ver, T., Marine-Font, A. & Vidal-Carou, M. (1996). Biogenic aminesin European beers. JAgric Food Chem 44, 3 1 59-3 1 63 .

55. Kamimote, S., Sumino, Y., Yamada, H., Imayasu, S.,Ichikawa, E. & Suizu, T. (1989). Quality improvementof alcoholic liquors. U.S. Patent 4.844.91 1.

56. Kellner, V. (1994). Most significant foreign substancesin brewing. Kvasny-Prumysl 40, 42-45.

57. Keraby, K. M. K., El-Sonbaty, A. H. & Badawi, R. M.(1999). Effects of heating milk and accelerating ripen-ing of low fat Ras cheese on biogenic amines and freeamino acids development. Food Chem 64, 67-75.

58. Kevekordes, S., Urban, M., Steffens, W., Janzowski,C., Eisenbrand, G. & Dunkelberg, H. (1996). Testingformation of nitrite and potency of nitrosation in alaboratory construction for heterotrophic biologicaldenitrification of drinking water. Zentralbl HygUmweltmedizin 198, 462-472.

59. Kim, S. H., An, H. & Price, R. J. (1999). Histamineformation and bacterial spoilage of albacore harvestedoff the U.S. Norwthest coast. J Food Sd 64, 340-343.

60. Kodame, S. & Yotsusuka, F. (1996). Acid urease: re-duction of ethyl carbamate formation in sherry undersimulated baking conditons. J Food Sd 61, 304-307.

61. Koh, E. & Kwon, H. (1996). Determination of fermen-tation specific carcinogen, ethyl carbamate, in kimchi.Korean J Food Sd Technol 28, 421-427.

62. Kozhev, S. & Kozhev, A. (1994). Potassium nitrateresidues in hard and semihard cheeses. Khranitelna-Promishlenost 43, 1 8-19.

63. Kunimote, M., Katsuhito, T. & Kaneniwa, M. (1992).Consumption of trimethylamine and trimethylamineN-oxide by Aspergillus oryzae. Nippon SuisanGakkaishiSZ, 1471-1476.

64. Lavanchy, P. & Sieber, R. (1993). Proteolysis in differ-ent hard and semihard cheeses. II. amines. SchweizMilchmrtsch Forsch 22, 65-68.

65. Lehtonen, P., Saarinen, M., Vesanto, M. & Riekkola,M. L. (1992). Determination of wine amines by HPLCusing automated precolumn derivation with o-ph-

thalaldehyde and fluorescence detection. Z LebensmUnters Forsch 194, 434^137.

66. Leuschner, R. G. K. & Hammes, W. P. (1998). Degra-dation of histamine and tyramine by Brevibacteriumlinens during surface ripening of Munster cheese. JFoodProt 61, 874-878.

67. Leuschner, R. G. K. & Hammes, W. P. (1998).Tyramine degradation by micrococci during ripeningof fermented sausage. Meat Sd 49, 289-296.

68. Leuschner, R. G. K., Kurihara, R. & Hammes, W. P.(1998). Effect of enhanced proteolysis on formation ofbiogenic amines by lactobacilli during Gouda cheeseripening. Int J Food Microbiol 44, 15-20.

69. Lonvaud-Funel, A. & Joyeux, A. (1994). Histamineproduction by wine lactic acid bacteria: isolation of ahistamine-producing strain of Leuconostoc oenos. JAppl Bacteriol 77, 401-407.

70. Lovaas, E. (1991). Antioxidative effects ofpolyamines. JAm Oil Chem Soc 68, 353-358.

71. Luten, J. B., Bouquet, W., Seuren, L. A. J., Burggraaf,M. M., Riekwel-Booy, G., Durand, P., Etienne, M.,Gouyou, J. P., Landrein, A., Ritchie, A., Leclerq, M. &Guinet, R (1992). Biogenic amines in fishery products:standardization methods within EC. In Quality Assur-ance in the Fish Industry. Denmark: Ministry of Fish-eries.

72. Maanen, J. M. S. van, Welle, I. J., Hageman, G.,Dallinga, J. W., Martens, P. L. J. M. & Kleinjans, J. C.S. (1996). Nitrate contamination of drinking water: re-lationship with HPRT variant frequency in lymphocyteDNA and urinary excretion of N-nitrosamines.Environ Health Perspect 104, 522-528.

73. Maijala, R. L., Eerola, S. H., Aho, M. A. & Him, J. A.(1993). The effect of GDL-induced pH decrease on theformation of biogenic amines in meat. J FoodProt 56,125-129.

74. Maijala, R., Eerola, S., Lievonen, S., Hill, P. & Hirvi, T.(1995). Formation of biogenic amines during ripening ofdry sausages as affected by starter culture and thawingtime of raw materials. J Food Sd 60, 1 1 87-1 190.

75. Massey, R. C. (1997). Estimation of daily intake offood preservatives. Food Chem 60, 177-185.

76. Masson, F., Lebert, A., Talon, R. & Montel, M. C.(1997). Effects of physico-chemical factors influenc-ing tyramine production by Carnobacteriumdiver gens. J Appl Microbiol 83, 36^7.

77. Masson, F. & Montel, M. C. (1995). Biogenic aminesin meat products. Viand Prod Carnes 16, 3-7.

78. Mirvish, S. S. & Sams, S. S. (1984). N-nitroso com-pounds: occurrence, biological effects and relevance tohuman cancer. In International Agency for Researchon Cancer, pp. 283-289. Edited by I. K. O'Neil, R. C.Von Borstel, C. T. Miller, J. Long & H. Bartsch. Lyon,France: Scientific Publications.

79. Mizutani, T., Kimizuka, A., Ruddle, K. & Ishige, N.(1992). Chemical components of fermented fish prod-ucts. J Food Comp Anal 5, 152-159.

80. Moret, S., Conte, L. S. & Spoto, E. (1996). Biogenicamines in cheese: parameters affecting their analyticaldetermination. Ind Aliment 35, 788-792.

8 1 . National Academy Sciences/National Research Coun-cil. (1981). The health effects of nitrate, nitrite and N-nitroso compounds. National Academy Press.

82. National Academy Sciences/National Research Coun-cil. (1982). Alternatives to the current use of nitrite infoods. National Academy Press.

83. Neamat-Allah, A. A. (1999). Biogenic amines inKarish and Mish cheese in Egypt. Egyptian J Dairy Sd25, 337-348.

84. Nguen, D. C. & Velikaya, E. (1972). Conditions forsaccharification and fermentation of manioc mash.Ferment Spirtov Promysh Lennost 1 , 1 7—2 1 .

85. Nout, M. J. R. (1994). Fermented foods and foodsafety. Food Res Int 27, 291-298.

86. Nout, M. J. R., Ruikes, M. M. W., Bouwmeester, H. M.& Beljaars, P. R. (1993). Effect of processing condi-tions on the formation of biogenic amines and ethylcarbamate in soybean tempe. JFoodSafll, 293-303.

87. Oesterdahl, B. G. & Alriksson, E. (1990). Volatile nit-rosamines in microwave-cooked bacon. Food Add andContaml, 51-54.

88. Offman, D., Djordjevic, M. V. & Brunnemann, K. D.(1991). New brands of oral snuffs. Food Chem Toxicol29, 65-68.

89. Ologhobo, A. D., Adegede, H. I. & Maduagiwu, E. N.(1996). Occurrence of nitrate, nitrite and volatile nitro-samines in certain feedstuffs and animal products. NutrHealth 11, 109-114.

90. Ough, C. S., Huang, Z. & Stevens, D. (1991). Aminoacid uptake by four commercial yeasts at two differenttemperatures of growth and fermentation: effects onurea excretion and reabsorption. Am J Enol Vitic 42,26-40.

91. Park, K. Y. (1995). The nutritional evaluation, and an-timutagenic and anticancer effects of kimchi. J KoreanSoc Food Nutr 24, 169-182.

92. Petridis, K. D. & Steinhart, H. (1996). Biogenic aminesin hard cheese production. I. factors influencing thebiogenic amine content in end-product by way ofSwiss cheese. Dtsch Lebensm Rundsch 92, 1 14-120.

93. Pfundstein, B., Tricker, A. R., Theobald, E.,Spiegelharder, B. & Preussmann, R. (1991). Maindaily intake of primary and secondary amines fromfoods and beverages in West Germany in 1989-1990.Food Chem Toxicol 29, 733-739.

94. Pogorzelski, E. (1992). Studies on the formation of his-tamine in must and wines from elderberry fruit. J SdAgric 60, 239-244.

95. Poocharoen, B., Barbur, F. J., Libbey, L. M. &Scanlan, R. A. (1992). Precursors of N-nitrosodimeth-ylamine in malted barley. I. determination of ordenineand gramine. J Agric Food Chem 40, 2216—2221.

96 . Potthast, K. ( 1 99 1 ) . Nitrite and nitrate — key compoundsfor a number of reactions during curing of meat andmeat products. Mitt Gebiete Lebensm Hyg 82, 24-35.

97. Pylypiw, H. M. Jr., Harrington, G. W., Kozeniauskas,R. & Diaz, D. (1992). The formation of N-nitrosodimethylamine and N-nitrodimethylamine fromthe plant growth regulator Daminozide (Alar). J FoodProt 55, 218-219.

98. Reichert, J. E. (1994). Influence of ascorbic acid andsodium ascorbate on quality of meat products. AlimentEquipos Tecnol 13, 67-68.

99. Rodriguez-Jerez, J. J., Lopez-Sabater, E. I.,Hernandez-Herrero, M. M. & Mora- Ventura, M. T.(1994). Histamine, putrescine and cadaverine forma-tion in Spanish semipreserved anchovies as affected bytime/temperature. J Food Sd 59, 993-997.

100. Roig-Sagues, A. & Eerola, S. (1997). Biogenic aminesin meat inoculated with Lactobacillus sake starterstrains and an amine-positive lactic acid bacterium.Food Res Technol 205, 227-23 1 .

101. Ross, H. D., Henion, J., Babish, J. G. & Hotchkiss, J.H. (1987). Nitrosating agents from the reaction be-tween methyl oleate and dinitrogen trioxide: identifica-tion and mutagenicity. Food Chem 23, 207-222.

102. Rudnitskii, P. V. (1973). New methods for improvingthe quality of alcohol and molasses by-products. TrudyUkrainskii Nauchno Inst Spirtovoi LikeroPromyshlnnosti 15, 5-9.

103. Sayem-El-Daher, N., Simard, R. E. & Pillion, J.(1984). Changes in the amine content of ground beefduring storage and processing. Lebensm Wiss Technol7,319-323.

104. Schneller, R., Good, P. & Jenny, M. (1997). Influenceof pasteurized milk, raw milk and different ripeningcultures on biogenic amine concentrations in semi-softcheeses during ripening. Food Res Technol 204, 265-272.

105. Sen, N. P., Kushkawa, S. C., Seaman, S. W. &Clarkson, S. G. (1985). Nitrosamines in baby bottlenipples and pacifiers: occurrence, migration, and effecton infant formulas and fruit juices on in vitro formationof nitrosamines under stimulated gastric conditions. /Agric Food Chem 33, 428^33.

106. Sen, N. P., Seaman, S. W., Bergeron, C. S. W. &Brousseau, R. (1996). Trends in the levels of N-nitrosodimethylamine in Canadian and imported beers.J Agric Food Chem 44, 1498-1501.

107. Shahidi, F. & Pegg, R. B. (1994). Absence of volatileN-nitrosamines in cooked nitrite-free cured musclefoods. Meat Sd 37, 327-336.

108. Shahidi, F., Synowiecki, J. & Sen, N. P. (1995). N-nit-rosamines in nitrite-cured chicken-seal salami. J FoodProt 58, 446^48.

109. Shalaby, A. R. (1993). Survey on biogenic amines inEgyptian foods: sausage. J Sd Food Agric 62, 291-293.

1 10. Shalaby, A. R. (1996). Significance of biogenic aminesto food safety and human health. Food Res Int 29, 675-690.

111. Shenoy, N. R. & Choughuley, A. S. U. (1989). Effectof certain plant phenolics on nitrosamine formation. JAgric Food Chem 37, 721-725.

112. Silla, M. H. (1985). Utilization de microorganismosfermentadores en Ia conservation de alimentos. RevAgroquim Tecnol Aliment 25, 170-182.

113. Silla-Santos, M. H. (1996). Biogenic amines: their im-portance in foods. Int J Food Microbiol 29, 2 1 3-23 1 .

1 14. Silla-Santos, M. H. (1998). Amino acid decarboxylasecapability of microrganisms isolated in Spanish fer-mented meat products. Int J Food Microbiol 39, 227-230.

115. Silva, C. C. G., Ponte, D. J. B. da & Dapkevicius, M. L.A. E. (1998). Storage temperature effect on histamineformation in big eye tuna and skipjack. J Food Sd 63,644-647.

116. Smith, N. A. (1994). Nitrate reduction and N-nitrosation in brewing. JInstBrew 100, 347-355.

117. Soloviev, A. L, Pivovarov, A. L, Knyazev, V. A. &Starushenko, A. K. H. (1987). Storage of sugar beetunder carbamide formaldehyde foamed plastics.Sakkarayc Promyshlennost 1, 40^41.

118. Soufleros, E., Barrios, M. L. & Bertrand, A. (1999).Correlation between the content of biogenic aminesand other wine components. Am J Enol Vitic 49, 266-278.

1 19. Sponholz, W. R., Kuerbel, H. & Dittrich, H. H. (1991).Formation of ethyl carbamate in wine. Wein Wissensch46,11-17.

120. Stevens, D. F. & Ough, C. S. (1993). Ethyl carbamateformation: reaction of urea and citrulline with ethanolin wine under low to normal temperature conditions.Am J Enol Vitic 44, 309-312.

121. Stratum, J. E., Hutkins, R. W. & Taylor, S. L. (1991).Biogenic amines in cheese and other fermented foods:a review. JFoodProt 54, 460^t70.

122. Straub, B. W., Kicherer, M., Schilcher, S. M. &Hammes, W. P. (1995). The formation of biogenicamines by fermentation organisms. Z LebensmUntersuch Forsch 201, 79-82.

123. Straub, B., Schollenberger, M., Kicherer, M., Luckas,B. & Hammes, W. P. (1993). Extraction and determi-nation of biogenic amines in fermented sausages andother meat products using reversed-phase HPLC.Lebensm. Untersuch Forsch 197,230-232.

124. Surono, I. S. & Hosono, A. (1996). Antimutagenicityof milk cultured with lactic acid bacteria from dadihagainst mutagenic terasi. Milchwissensch 5 1 , 493^97.

125. Teodorovic, V. & Buncic, S. (1999). The effect of pHon tyramine content in fermented sausages.Fleischwirtsch 79, 85-88.

126. Teodorovic, V. B., Buncic, S. & Smiljanic, D. (1994).A study of factors influencing histamine production inmeat. Fleistwirtsch 74, 181-183.

127. Tiecco, G., Tantillo, G., Francioso, E., Paparella, A. &Natale, D. (1986). Qualitative and quantitative deter-mination of some biogenic amines in sausages duringripening. Ind Aliment 25, 209-213.

128. Tricker, A. R. & Preussmann, R. (1990). Chemicalfood contaminants in the initiation of cancer. SympDiet Cancer Proc Nutr Soc 49, 133-144.

129. Trova, C., Gandolfo, G. & Cossa, G. (1993). N-nitro-sodimethylamine levels in beer. IndBev 22, 528-530.

130. Tsen, C. C. (1973). Chemical dough development.Bakers ' Digest 47, 44^47, 129.

131. Tunon, H. (1994). Hops — not just for adding bitternessto beer. Kem Tidskrift 106, 30-32.

132. Vale, S. R. & Gloria, M. B. A. (1997). Determinationof biogenic amines in cheese. J AOAC Int 80, 1006-1012.

133. Valeiro Novo, C. & Pilas Perez, A. (1994). Problemassanitarios asociados al consumo de pescados. Aliment250, 55-67.

134. Veciana-Nogues, M. T., Marine-Font, A. & Vidal-Carou, M. C. (1997). Biogenic amines as hygienicquality indicators of tuna: relationships with microbialcounts, ATP-related compounds, volatile amines, andorganoleptic changes. J Agric Food Chem 45, 2036-2041.

135. Vidal-Carou, M. C., Izquierdo-Pulido, M. L., Martin-Morro, M. C. & Marine-Font, A. (1990). Histamineand tyramine in meat products: relationship with meatspoilage. Food Chem 37, 239-249.

136. Vincze, L. (1979). Factors affecting feed protein utili-zation by broiler chickens. Agrar Koezl 38, 393-398.

137. Watkins, S. J., Hogg, T. A. & Lewis, M. J. (1996). Theuse of yeast inoculation in fermentation for port pro-duction; effect on total potential ethyl carbamate.Biotechnol Lett 18, 95-98.

138. Wendakoon, C. N. & Sakaguchi, M. (1993). Combinedeffect of sodium chloride and clove on growth and bio-genic amine formation of Enterobacter aerogenes inMackerel muscle extract. JFoodProt 56, 410—413.

139. Yankah, V. V., Ohshima, T. & Koizumi, C. (1993).Effects of processing and storage on some chemicalcharacteristics and lipid composition of a Ghanaianfermented fish product. JSd Food Agric 63, 227-235.

140. Yatsunami, K. & Echigo, T. (1993). Studies onhalotolerant and halophilic histamine-forming bacte-ria. III. Changes in the number of halotolerant hista-mine-forming bacteria and contents of non-volatileamines in sardine meat with addition of NaCl. Bull JpnSoc Sd Fish 59, 123-127.

141. Yen, G. C. & Kao, H. H. (1993). Antioxidative effectof biogenic amine on the peroxidation of linoleic acid.Biosci Biotech Biochem 57, 115-116.

CHAPTER 7

Microbiological Hazards andTheir Control: Bacteria

R. R. Beumer

INTRODUCTION

The inactivation or inhibition of undesirablemicroorganisms is an essential part of food pres-ervation. Foods can be made safe by treatmentssuch as pasteurization or sterilization (e.g., milk,cooked meat products, and canned foods), orthey may possess intrinsic properties that con-tribute to safety, such as structure, low pH, and/or low water activity (Aw) (e.g., fruit, pickles,and honey). One further way of improving shelflife and safety, through the use of a competitivemicrobial flora, is widely employed in the pro-duction of fermented milks, vegetables, meat,fish, and grains, giving products with changedcomposition and taste as well as a prolongedshelf life. It is estimated that approximately 30%of our food supply is based on fermented prod-ucts,40 though not all fermented foods are uni-versally accepted. The modified sensory proper-ties in some fermentation products areacceptable only in certain regions; elsewhere,they would be considered spoiled or unpalatable.

Fermented foods have generally been consid-ered as less likely to be vehicles for food-borneinfection or intoxication than fresh foods due tothe competitive activity and metabolites of thefunctional flora49 (see Chapter 2). All raw mate-rials used in fermentation processes, such asfruits, vegetables, meat, fish, milk, and eggs, canharbor pathogenic bacteria.60 Depending on thetype of fermentation (e.g., lactic acid, alcoholic,or mold), these pathogens may be (partly) elimi-nated or their growth inhibited by the antibacte-

rial substances that are produced (e.g., lacticacid, acetic acid, alcohol). To ensure a satisfac-tory fermentation and to control the growth offood-borne pathogens, the use of specific start-ers is often recommended.

In this chapter, the safety of fermented productswill be discussed in relation to the pathogenic bac-teria: Aeromonas hydrophila, Bacillus cereus,Campylobacter, Clostridium botulinum, Escheri-chia coll O157:H7, Listeria monocytogenes, Sal-monella spp., Shigella, Staphylococcus aureus,Vibrio spp., and Yersinia enter ocolitica.

OUTBREAKS RELATED TOFERMENTED PRODUCTS

A great part of food-borne disease is prevent-able. In addition to good manufacturing prac-tices (GMPs) and the Hazard Analysis CriticalControl Point (HACCP) system, domestic hy-gienic practices have a key role to play in pre-venting infection.9'12'52 Most food-borne diseasesare attributed to the consumption of, or cross-contamination from, raw or undercooked foodproducts (e.g., vegetables, sprouts, shellfish, andpoultry), in many cases with an extended periodbetween preparation and consumption.14

The antibacterial factors that are present infermented foods may affect both the growth andthe survival of bacterial pathogens that arepresent in a raw material. In most fermentedfoods, the inhibition of growth is more commonand can often ensure safety where levels of con-tamination are low. But with infectious patho-

gens, particularly those with a small minimuminfectious dose, some degree of inactivation maybe necessary to provide an acceptable level ofsafety. So, although fermented foods are gener-ally considered to be safe, process failures andcontaminated raw materials have resulted intheir being involved in food-borne illness. Forexample, several outbreaks of illness have beenattributed to the consumption of fermented sau-sages contaminated with S. aureus and Salmo-nella spp.,70 and emerging pathogens, such as L.monocytogenes and E. coll O157:H7, have beenidentified as causative organisms in outbreaksinvolving fermented products such as sausages,cheeses, and yogurt.8'27'47

These outbreaks have raised questions regard-ing the safety of fermented food products andhave prompted studies using artificially con-

taminated raw materials to identify critical con-trol points (e.g., fermentation pH, final heatingtemperature, heating time) and critical limits(e.g., pH 4.6, 55 0C for 20 minutes) that can beintegrated into an HACCP plan to ensure safeproducts.23

BACTERIAL PATHOGENS OFCONCERN IN FERMENTEDPRODUCTS

There is extensive literature describing thelimits for the growth of bacterial pathogens.Some of the data relevant to fermentation pro-cesses are summarized in Table 7-1, 61 and a fewkey features of these organisms are presented inthe following paragraphs.

Table 7-1 Limits for Growth of Some Common Bacterial Pathogens

Minimum AwOrganism Minimum pH Minimum Temperature (Max % NaCI)

Aeromonas hydrophila < 4.5 > 0.0, < 4.0 -(5-6%)

Bacillus cereus 5.0 4.0 0.93Campylobacter 4.9 30.0 > 0.987

(1.5%)Clostridium botulinum

Group 1: Proteolytic 4.6 10.0-12.0Group 2: Nonproteolytic 5.0 3.3 (10%)

(5%)Escherichia coli 4.4 7.0-8.0 0.95Listeria monocytogenes 4.4 -0.4 0.92Salmonella 3.8 5.2* 0.94Shigella 4.9-5.0 6.1-7.9

(3.78-5.18%)Staphylococcus aureus

Growth 4.0 7.0 0.83Toxin production 4.5 10.0 0.87

Vibrio cholerae 5.0 10.0 0.97(4.0%)

Vibrio parahemolyticus 4.8 5.0 0.94(10%)

Yersinia enterocolitica 4.2 -1.3(> 5%,< 7%)

*Most serotypes fail to grow below 7 0C.

Aeromonas

Aeromonas are facultatively anaerobic, Gram-negative rods (belonging to the Vibrionaceae)that are found in water (including sewage) and infood products that have been in contact withcontaminated water (e.g., seafood, vegetables).Some species are well known as fish pathogens;others are human pathogens. A. hydrophilaseems the most important species; however, itssignificance in the epidemiology of human ill-ness is still unclear. Symptoms associated withinfection include diarrhea, abdominal pain, andheadache, and are more severe in children.58'61

Aeromonas do not appear to be a serious riskin well-produced fermented foods. The additionof Aeromonas to skim milk during lactic fermen-tation,57 to yogurt,6 and to two traditional fer-mented foods (mahewu and sour porridge) re-sulted in a sharp decrease in numbers.65

However, microbiological examination ofhomemade fermented sorghum porridge41 andindustrially produced fermented sausages(longaniza and chorizo) showed its presence innumbers ranging from 1 .0 logio cfu/g to 4.5 logiocfu/g. It was concluded that the hygienic statusof factories significantly affected the incidenceand counts of Aeromonas.24

Campylobacter

Campylobacter species are micro-aerophilic,Gram-negative curved or spiral rods belongingto the Campy lobacteriaceae. Members of thisgenus have been associated with disease in ani-mals for many years. From 1970, the thermo-philic campylobacters (growth from 30^45 0C)have been recognized as important food-bornepathogens; they are now the most common causeof gastroenteritis in many developed countries.C. jejuni, C. coll, and C. lari are responsible formore than 90% of the human illness caused bycampylobacters, which are normally character-ized by profuse diarrhea and abdominal painthree to five days after consumption. Diarrheacan often be bloody, and more serious sequelaesuch as reactive arthritis and Guillain-Barre syn-drome have also been reported.1

Enormous numbers of campylobacters maybe found in the gut of chickens (and other birds)and pigs. C. lari is often found in seagulls and, asa consequence, in shellfish.34'61 Raw milk andpoorly cooked poultry products are most fre-quently involved in outbreaks, and cross-con-tamination from raw poultry to ready-to-eatproducts may also be an important factor intransmission. Although the infective dose ofCampylobacter is relatively low (approximately500 cfu), Campylobacter transmission by fer-mented foods is probably not significant. Theydo not grow below 30 0C and are sensitive tofreezing, drying, low pH, and sodium chloride,and so would not survive well in fermentedfoods. This has been confirmed by artificial con-tamination of fermenting products such assalami, weaning foods, and yogurt, which re-sulted in a rapid decrease in numbers.39'44'48

Vibrio spp.

Vibrios are facultatively anaerobic, Gram-negative bacteria belonging to the family of theVibrionaceae, and are commonly found inaquatic environments.

V. cholerae includes the true cholera-producingbacteria ("classic" and "El Tor" biovars) and theso-called "non 0 1 strains," which lack the somatic"Ol" antigen and produce a less severe gastroen-teritis. Cholera has been known since ancienttimes and is still a disease of worldwide signifi-cance. The organism is commonly isolated fromriver waters and coastal marine waters and, as aconsequence, from shellfish and other marine ani-mals,61'71 but its transmission by fermented fishproducts has not been reported. The use of con-taminated water for washing or irrigation may in-troduce the pathogen onto vegetables. Refriger-ated or frozen storage of contaminated foodproducts will not ensure safety because of thegood survival of vibrios at low temperatures.19

V. parahaemolyticus is a halotolerant patho-gen that is commonly found in marine coastalwaters in warm water regions. Fish and shellfishare usually associated with outbreaks caused bythis pathogen, as are, to a lesser extent, salt-pre-served vegetables.18' 61

V. vulnificus is very similar to V. parahaemo-lyticus but differs from the other pathogenic,halotolerant vibrios in that diarrhea is not a com-mon symptom of infection. V. vulnificus ishighly invasive, causing septicemia. Illness isclearly associated with liver diseases such as cir-rhosis and hepatitis. The majority of cases of ill-ness have been reported after the consumption ofraw shellfish, in particular, raw oysters.38'61

Vibrios are not acid tolerant and lactic acidbacteria (LAB) isolated from fermented fishproducts inhibited both V. cholerae and V.parahaemolyticus, indicating that the antibacte-rial activity of the functional flora is important inensuring the safety of fermented fish products.56

This has also been observed in pickled foods andsquid shiokara that were artificially contami-nated with V. parahaemolyticus.12^

E. coli O157:H7, Salmonella spp., Shigellaspp., and Y. enterocolitica

All members of the Enterobacteriaceae familyare facultatively anaerobic, oxidase-negative,Gram-negative rods that are capable of ferment-ing glucose. Some are known pathogens, such asE. coli (especially E. coli 0 157), Salmonella,Shigella, and Yersinia, and a great part of thefamily, including the food-borne pathogens, arecomponents of the fecal flora of animals. As aresult, the whole family or parts of it (coliformsand E. coli) are used as hygiene indicators forprocessed foods.

E. coli O157:H7

On the basis of virulence factors and diseasepatterns, human pathogenic E. coli strains areclassified in the following principal groups:

• Enteropathogenic E. coli (EPEC), acutewatery diarrhea; young children particu-larly susceptible

• Enterotoxigenic E. coli (ETEC), acute wa-tery diarrhea (often in travelers)

• Entero-invasive E. coli (EIEC), dysentery-like syndrome

• Enterohemorrhagic E. coli (EHEC), bloodydiarrhea syndrome

Within the EHEC-group, E. coli O157:H7 isthe organism that is isolated most frequently incases of hemorrhagic colitis (HC) and hemolyticuremic syndrome (HUS). Because this pathogenis relatively acid resistant, it is less affected byfermentation and can be of concern in fermentedfoods such as yogurt and fermented sau-sages.22'25'43'61 Several food-borne outbreaks ofE. coli have been reported in recent years, in-cluding some involving fermented meats andcheeses.17'63'64'67'69

Salmonella

Salmonella is often found on raw meat (poul-try) and is widely distributed in the environmentand universally recognized as an importantcause of food-borne infections. Salmonellas arepresent in the gut of contaminated animals (andhumans) and are shed in the feces. As a result, agreat variety of raw materials can be contami-nated with this pathogen. Illness is characterizedby diarrhea, abdominal pain, and fever generally6-48 hours after consumption of the organism.1

Meat, milk, poultry, and eggs are the principlevehicles for Salmonella transmission, and fer-mented foods derived from these materials, suchas salami and cheeses, have occasionally beenassociated with outbreaks of illness. Recently,acid-resistant strains of Salmonella have beenimplicated in food-borne outbreaks,42'61 whichled to many studies on the fate of salmonellasinoculated in fermenting products such as fingermillet,4 Siljo,21 and fermented sausages.23'33'47'62

Shigella

Shigella is closely related to E. coli (DNA ho-mology and biochemical pattern). Shigellas arenot natural inhabitants of the environment butoriginate from humans and higher primates inthe acute phase of illness. Both person-to-personinfections (due to poor personal hygiene) and theconsumption of contaminated water or foodswashed with contaminated water give rise to dis-tinct symptoms of the disease, such as bloody

diarrhea. In some countries, shigellas are epi-demic.61 Artificial contamination of productswith low pH values (< 5.0) results in a rapid de-crease in numbers of shigellas, suggesting thatmost acid-fermented foods will not be signifi-cant as vehicles for this type of infection.41'53

Y. enterocolitica

7. enterocolitica is a widely distributedpsychrotrophic bacterium. Certain serotypes(O:3, 0:5, 27; O:8, and O:9) present in raw milk,seafood, and raw pork have been associated withfood-borne infections. The tonsil area of pigs is aunique ecological niche with a high incidence ofY. enterocolitica. Gastroenteritis is the mostcommon symptom; however, sometimes illnessis also characterized by fever and abdominalpain (resembling appendicitis). In some cases,this has led to surgery.61 There are relatively fewoutbreaks reported with this pathogen; however,in many of these, milk was the incriminatedproduct. The behavior of Yersinia has been stud-ied in fermenting milk57 and yogurt. 1 1 During thefirst few hours of the fermentation, growth wasobserved, followed by a reduction below the de-tection level after fermentation and storage forfour days.

B. cereus

B. cereus is an aerobic, Gram-positive spore-forming rod (belonging to the Bacillaceae) thatcan cause both food-borne infection (character-ized by diarrhea) and intoxication (vomiting af-ter ingestion of the toxin cereulide). As a sporeformer, this bacterium is ubiquitous and may befound in cereals and their products, such as riceand flour, and spices. Small numbers in foodsare not considered significant.61 The behavior ofB. cereus was studied in fermenting productssuch as mageu, a sour maize beverage;15 kinema,a fermented soya bean food;50 tempeh;51 salads;13

and fish sausage.5 In all cases, initial growth wasobserved, but in products where pH decreaseddue to the growth of LAB, subsequent inhibitionoccurred that was correlated to the rate of de-crease in pH. In products without a lactic acid

fermentation (e.g., tempeh), numbers reached108 cfu/g. However, if the soya beans weresoaked, resulting in acidification of the beans topH less than 4.5, the growth of B, cereus wasprevented.51 Interestingly, food-borne illness hasalso been associated with other Bacillus speciessuch as B. subtilis, B. licheniformis, and B.pumilis, and strains of many of these are associ-ated with the production of fermented vegetableprotein products such as Japanese natto, Nige-rian iru (dawadawa), Indian kinema, and Thaithua-nao, but, as far as we are aware, there havenot been any recorded outbreaks of illness asso-ciated with them.

C. botulinum

C. botulinum is an anaerobic, Gram-positive,spore-forming rod (belonging to the Bacil-laceae). Food-borne botulism is a neuroparalyticsyndrome caused by the botulinum neurotoxinthat is produced by vegetative cells of C. botuli-num growing in food. This bacterium is ubiqui-tous, and spores are widely distributed in the soiland on the shores and bottom deposits of lakesand coastal waters, and in the intestinal tract offish and animals.61 The human pathogens areclassified into proteolytic and nonproteolyticstrains; the former being slightly more acid tol-erant and mesophilic, whereas the latter cangrow down to 3.3 0C. Most outbreaks of food-borne botulism have been caused by home-pro-cessed vegetables, fish, or meat products.59 Al-though C. botulinum spores are present in mostraw agricultural products, the number of out-breaks is relatively small, usually occurring afterinadequate processing. Nitrite inhibits C. botuli-num and is an important safety factor in the pro-duction of cured and fermented meats. In Japan,botulism caused by nonproteolytic strains pro-ducing type E toxin is often associated with theconsumption of izushi, which is made by thenatural fermentation of raw fish and cookedrice.32 The largest recorded outbreak of food-borne botulism in the United Kingdom was asso-ciated with the consumption of contaminatedhazelnut yogurt, but in this case, production of

the toxin had occurred in the hazelnut puree thatwas added to the yogurt rather than in the yogurtitself (see also Chapter 5).55

S. aureus

S. aureus is a facultatively anaerobic, Gram-positive coccus belonging to the Micrococ-caceae. It can produce several heat-stable en-terotoxins that, after ingestion of the food, maycause food poisoning with symptoms of nauseaand vomiting. The production of staphylococcalenterotoxins occurs mostly in food products thatare rich in proteins with a low number of com-petitive microorganisms (e.g., whipped cream,cooked meat products, smoked fish). S. aureus ispresent on the skin and mucous membranes ofwarm-blooded animals (including humans). Ap-proximately 50% of the human population arecarriers of this pathogen. Contamination ofcooked or ready-to-eat food products by a colo-nized person followed by storage for severalhours at ambient temperatures is often impli-cated in outbreaks of illness.61 The organism isrelatively salt tolerant and will produce toxindown to pH values of 4.5. Fermented sausages7

and, to a lesser extent, raw milk cheeses35'36 havebeen associated with outbreaks of illness, al-though the organism is generally regarded as apoor competitor and its growth in fermentedfoods is generally associated with a failure of thenormal flora.54

L. monocytogenes

L monocytogenes is a facultatively anaerobic,Gram-positive rod. Species ofListeria are ubiq-uitous and potential food contaminants, but onlyL monocytogenes is a human and animal patho-gen. Approximately 30% of cases of listeriosisare perinatal, and in 20-25% of the cases, infec-tion proves fatal for the fetus or newborn. Mostother cases of listeriosis occur in immunocom-promised persons, with a death rate varyingfrom 30% to 50%.61

Due to its widespread occurrence in the envi-ronment, this pathogen can be harbored in lownumbers by practically all raw food products,

and processed foods can be contaminated fromthe production environment. L. monocytogenesis able to grow at refrigeration temperatures, so aconsiderable increase in numbers is possible incontaminated products with a long shelf life(e.g., cooked meat products, smoked salmon).

Fermented products such as sausages and softmold-ripened cheeses have been associated withlisteriosis.26'30 The organism is not particularlyacid tolerant, but where mold ripening occurs ina fermented product, the rise in pH can allowsurviving Listeria to resume growth, causingproblems to occur.

PRESENCE OF PATHOGENS IN RAWMATERIALS

Bacteria can enter raw materials or food prod-ucts at various stages in the food chain. Primaryproduction is a major source of both spoilageand pathogenic organisms. The primary con-tamination of raw materials includes microor-ganisms from soil, feces, surface water, and soforth, and will also be influenced by conditionsof harvesting, slaughter, storage, and transport.During food production, the extent of secondarycontamination will be influenced by factory lay-out, hygienic design of the equipment, personalhygiene, water, air, (pest) animals, and packag-ing material.10

Many of the bacteria introduced into food areof little concern. They are not well adapted tosurvival or growth in the product or are easilyeliminated in further processing steps. However,a certain proportion of bacteria may cause prob-lems and are a good reason to try to keep the mi-crobial load of raw materials as low as possible,particularly for those products that reach theconsumer in a raw state (e.g., meat, poultry, sea-food, eggs, and vegetables). Pathogen-free rais-ing of livestock would reduce risk, but this is dif-ficult and expensive to achieve. There is clearlyan urgent need for efficient and acceptable meth-ods to decrease the numbers of spoilage micro-organisms and pathogens from such products.Ionizing radiation is very effective, but con-sumer acceptance of this procedure is still a

problem. Decontamination by a lactic acid dip orspray decreases the numbers of some pathogenson carcasses, although acid sprays appear to pro-duce little reduction in E. coll (including E. coll0 157) and Salmonella.2^66 Food legislators inmany countries also have objections to thismethod on the grounds that it may mask poorhygiene or lead to the adaptation of microorgan-isms to low pH.

Table 7-2 presents the relationship betweenthe pathogens discussed in this chapter and rawmaterials used in fermentation processes.

It has been estimated that approximately 30%of the human food supply is based on fermentedproducts. Raw materials used are vegetables(e.g., cabbage, cucumber, olives), cereals (e.g.,rice, maize, wheat), legumes (e.g., soybean),fruit (e.g., grapes), meat (particularly pork), andmilk.16 From Table 7-2, it is clear that these rawmaterials may be contaminated with usually lownumbers of some of the pathogens listed.

Inactivation, survival, and/or growth during andafter fermentation depends on intrinsic factors(e.g., pH and Aw), extrinsic factors (e.g., storagetemperature and modified air packaging), implicitfactors (e.g., antagonism and the production of an-tibacterial substances), and process factors (e.g.,heat treatment), as well as the physiological char-acteristics of the organisms themselves.

EFFECT OF FERMENTATIONPROCESSES ON THE SURVIVAL OFBACTERIAL PATHOGENS

The effect of fermentation processes on thesurvival of bacterial pathogens is discussedmore extensively in Chapter 2. Before the mainfermentation process starts, nonmicrobial pro-cess factors such as peeling, washing, soaking,and/or grinding may reduce the numbers ofpathogens present in food. Removal of the outerlayer (peeling) has the greatest effect, whereaswashing of the raw materials usually reduces to-tal counts by 1 log cycle. Sometimes, a higherreduction can be obtained by adding (organic)acids or disinfectants; however, this procedure isnot allowed in all countries. Moreover, a majordisadvantage to adding acids or disinfectants isthe possibility of selecting strains adapted to lowpH or to disinfectants if concentrations are toolow to be detrimental.37'66'68 In particular, theacid-adapted pathogens have shown better sur-vival in fermented products.28'42

Processes such as mixing and grinding haveno lethal effect on bacteria. The only result ofthese steps is a better distribution of cells in theproduct and the separation of clumps of cells.

The main factors controlling the survival and/or growth of food-borne pathogens are (low) pH,

Table 7-2 Raw Materials and the Presence of Some Food-borne Pathogens

Pathogen Meat Milk Vegetables & Fruit Fish & Shellfish

Aeromonas ++ + +++ ++Bacillus cereus - ++ ++ -Campylobacter +++ +Clostridium botulinum ++ + ++ ++EscherichiacoliO157:H7 ++ ++ +Llsterla monocytogenes + + + +Salmonella ++ + + +Shigella - - + +Staphylococcus aureus ++ ++ -Vibrio - - + +++Yersinia + + - -

+ = sometimes present (25%), ++ = likely to be present (25-50%), +++ = usually present (> 50%), and - = not likely to be foundin these products.

Aw, heat treatment, and cold storage (see Chap-ter 2). In Table 7-3, the effect of various fermen-tation processes is shown on pH and Aw valuesin the final products and the possible productionof specific antimicrobial substances.

It is generally accepted that in well-fermentedfoods, the production of organic acids and a lowAw (due to the addition of salt or drying) maycontrol the growth of pathogens in these prod-ucts. However, in view of the values of pH andAw in the end products (Table 7-3), it can be con-cluded that many food-borne pathogens, espe-cially in products with relatively high final pHvalues, can survive. They may even grow if theAw is also high enough. A critical point for thegrowth of pathogens is the beginning of the fer-mentation process, when the pH is still high andthe competitive effect of the functional flora islow. Some alcoholic fermentations may be apossible exception, where a low substrate pHvalue may prevent pathogen growth.2'46

The abundance of literature on the inhibition offood-borne pathogens by LAB in fermentationprocesses is confusing. There are variations in thetypes of products (e.g., a wide variety of sausages,fermented milks, etc.) and considerable differ-ences in protocols for growth and addition of thepathogen, which makes comparison of the resultsobtained by different researchers very difficult.

During fermentation, antimicrobial productsare produced (see Chapter 2). Most of these, in-cluding organic acids, hydrogen peroxide, carbondioxide, diacetyl, and bacteriocins, are producedby LAB. There is considerable evidence on theantimicrobial activity of organic acids such as lac-tic acid and acetic acid, but effective concentra-tions of hydrogen peroxide, carbon dioxide, anddiacetyl are rarely present at levels sufficient tomake a major contribution to antibacterial activity.

The value of bacteriocins is limited by the factthat they are usually only active against Gram-positive cells, although theoretically this couldbe overcome by the addition of chelating agentssuch as Ethylenediaminetetraaceticacid, orEDTA (see Chapter 2). In practice, it might be ofmore advantage to concentrate on other factorssuch as choice of raw materials and the use ofGMP and the HACCP concept.

THE CONTROL OF MICROBIALHAZARDS

Although many fermented foods are a lesslikely vehicle for bacterial food-borne illnessthan fresh foods, hygienic practices during fer-mentation (especially in traditional householdfermentations), the possible use of contaminatedraw materials, and the low stability of these

Table 7-3 Effect of Various Fermentation Processes on the Survival and Growth of Food-bornePathogens

Type of FermentationSubstrates pH Aw Antagonism*

Lactic acid fermentationMeat, pork, poultry 4.5-6.0 0.85-0.99 B, H, C, DCabbage, vegetables < 4.0-5.5 > 0.96 B, H, C, DMilk < 4.0-5.5 > 0.96 B, H, C, D

Alcoholic fermentationGrains 4.0-5.5 >0.96 EFruits <4.0 >0.96 E

Mold fermentationSoybeans 4.0-5.Qt >0.96

*By production of other substances than organic acids (e.g., bacteriocins (B), hydrogen peroxide (H), ethanol (E), carbondioxide (C), diacetyl (D), etc.).

tpH value after soaking, just before inoculation with spores

products compared with canned or frozen foods,make control of microbial hazards necessary.The following factors are of importance:

• use of contaminated raw materials• prevention of contamination (zoning,

cleaning, and disinfection)• poorly controlled fermentations, including

ripening (HACCP)• consumption without prior heating

The Use of Contaminated Raw Materials

The prevention of contamination of raw mate-rials with pathogenic microorganisms should bethe goal of everyone involved in the preharvestand postharvest phases of delivering products tothe consumer. This is a "mission impossible"because pathogens are normally present in thesoil (and therefore on the surface of fruits andvegetables), in surface water (which results incontaminated fish and shellfish), and in the gutof animals (causing contaminated products ofanimal origin, such as milk and meat).

To minimize the risk of infection or intoxica-tion, pretreatment of the raw materials might behelpful. The simple practice of washing, for ex-ample, removes a portion of the pathogens (up to90%). Another possibility (not for raw meat) isheat treatment (pasteurization) of the raw mate-rials (e.g., milk). If pasteurization of the rawmaterials is not possible, a postprocess pasteur-ization might be necessary to eliminate patho-gens in the final product. This type ofpostprocess pasteurization is sometimes prac-ticed with fermented sausages.

Zoning

Raw materials are the primary source of manyof the bacteria that are responsible for food-borne infections and intoxications. But, bacteriamay also be transferred to food by the produc-tion environment and personnel, either directlyor by cross-contamination through surfaces,equipment, utensils, and/or hands that have notbeen adequately cleaned and disinfected. Zon-ing, that is, dividing the production area into

"dry and wet" and/or "high, medium, and lowcare" areas, is useful in preventing product con-tamination. The concept of hygiene zones wasinitially applied to prevent Salmonella contami-nation of very sensitive products such as milkpowder. Nowadays, zoning has evolved into acomplex set of measures including the layoutand design of equipment, air filtration, personnelhygiene routes, and appropriate cleaning anddisinfection procedures. However, zoning isonly useful when it is applied logically and to theappropriate degree.

Applied in fermentation processes, thereshould be areas for

• the storage of the raw materials (low hy-giene)

• the preparation of the raw materials, that is,washing, cutting, and adding ingredients(medium hygiene)

• the fermentation of the raw materials (me-dium hygiene)

• the filling of suitable packages (mediumhygiene)

• the storage of the final products (mediumhygiene)

An example of an area of high hygiene is theroom where starter cultures are prepared for thelactic fermentation of milk. Here, a potentialdanger is the infection of the cultures with bacte-riophages. Zoning, the allocation of an areastrictly separated from the production process,will help to prevent the introduction of bacte-riophages in starter cultures. However, this willbe insufficient if it is not accompanied by otherpreventive measures such as strict cleaning anddisinfection procedures and changing shoeswhen entering the area. Changing shoes cannotbe replaced efficiently by covering shoes withplastic covers (that often break) or by using shoedisinfecting systems (improper disinfection dueto rapid inactivation of the disinfectant), whichcan cause potentially contaminated wet areasaround the system.

It is not feasible to discuss zoning for all typesof fermentation processes here. In practice, zon-ing should be embedded in HACCP studies andmust be acceptable and practicable for all indi-

viduals. Zoning should be introduced after ad-equate training so that personnel know why zon-ing is important and are motivated to follow itsrequirements.

Cleaning and Disinfection

Inadequate cleaning and disinfection can leadto reduced food quality, resulting in more rapidfood spoilage and greater risk of food-borne dis-eases. Moreover, in food spills (pathogenic) mi-croorganisms might adapt to extreme conditionssuch as low pH, giving them the opportunity tosurvive and/or grow in the final product.

Cleaning and disinfection are two separate butclosely related concepts. Cleaning is removingdirt and a proportion of the microorganismspresent; disinfection is treating the surfaces insuch a way that the remaining microorganismsare killed or reduced to an acceptable level.Cleaning should be done first; otherwise, thesubsequent disinfection will be less effective.

HACCPs

After zoning has been introduced as part ofthe prevention plan, attention should be paid tothe presence and behavior of pathogens in rawmaterials, the fermentation process, the finalproduct, and the production environment. Mi-crobiological investigation estimating the pres-ence and numbers of pathogens and spoilage or-ganisms in raw materials and food productsforms an important part of control programs forensuring the safety and quality of food products.Until recently, microbial counts were an impor-tant though ineffective method for assessingfood quality. Nowadays, food safety systemshave been introduced based on HACCPs, inwhich control is exercised throughout the pro-cess itself.

An HACCP study will provide informationconcerning critical control points (CCPs), that is,raw materials, locations, practices, procedures,formulations, or processes where measures can beapplied to prevent or minimize hazards.45

It is outside the scope of this contribution toidentify all CCPs for known fermentation pro-cesses. In general, raw materials (contaminatedwith pathogens), the fermentation process (timeto reach final pH, Aw, and final pH), storage andpreparation of starter cultures, and heat treat-ment (of raw materials or of the fermented prod-ucts) are the most important CCPs.

An HACCP study will not result in controlmeasures that eliminate all safety problems, butit will provide information that can be used todetermine how best to control the remaininghazards. It is then up to management to use thatinformation correctly. If control procedures fol-low clearly established rules, inspectors canhave greater confidence in food producers. Gov-ernments are also more able to accept the re-sponsibility taken by industry because they canunderstand why and how controls are made.This has been recognized internationally, andthe application of HACCP principles is recom-mended by the Codex Alimentarius Commis-sion and is mandatory in many countries. In itspresent format, the application of HACCP prin-ciples is best suited to industrial food process-ing; small-scale producers of fermented foods,especially in nondeveloped countries, will havegreater difficulty following this route. Relativelysimple hygiene codes, based on HACCPs, couldinform food handlers and households about ap-propriate fermentation techniques46'52 (see Chap-ters 3 and 12).

Quantitative Risk Assessment

In many cases, the concept of HACCPs isused in only a qualitative way. By the implemen-tation of quantitative risk analysis (QRA) in ex-isting HACCP systems, a more quantitative ap-proach is possible. A smart, stepwise, interactiveidentification procedure for food-borne micro-bial hazards has been developed by van Gerwen& Zwietering.29 This procedure is based on threelevels of detail ranging from rough hazard iden-tification via detailed to comprehensive hazardidentification. At first, the most relevant prob-lems are identified before focusing on less im-portant problems.

• The first level (rough hazard identification)selects pathogens that were involved infood-borne outbreaks with the product inthe past. These data can be found in the lit-erature, although not all outbreaks are re-ported, particularly in nondeveloped coun-tries, and in many cases or outbreaks, thecausative organism or the food vehicle isnot identified.

• The second level (detailed hazard identifi-cation) selects pathogens that were re-ported to be present on the raw materialsand ingredients used in the product. Thegreater part of these data can be found inthe literature.

• The last level (comprehensive hazard identi-fication) selects all (human) pathogens ashazardous, including pathogens that unex-pectedly recontaminate the product. In thisstep, it is possible to estimate the effect ofunexpected hazards (i.e., emerging patho-gens, or pathogens that adapted to or are re-sistant to intrinsic factors of the product, suchas acid-resistant Salmonella). Possible futureproblems can be anticipated in this way.

After this type of hazard identification,knowledge rules should be used to reduce theprobably long list to a manageable list of thepathogens that are most likely present in the fi-nal product.

The following three types of rules are used.

1 . Selection of pathogens that are present orable to survive in the end product (e.g., re-moval from the list of vegetative patho-gens present in a product that will un-dergo pasteurization). Survival afterpasteurization due to inappropriate time/temperature combinations and post-process contamination are not included inthis rule. This should be ascertained byGMP and HACCP analysis.

2. Selection of pathogens that are likely tocause problems. Pathogens that are veryrarely transmitted are not likely to causehealth problems and can be removed fromthe list.

3. Selection of pathogens that are able togrow or produce toxin in the product. Esti-mation of capacity to grow is based on themaximum and minimum growth tempera-ture, pH, and Aw. Other growth-determin-ing factors such as the presence of preser-vatives and natural antimicrobial systemsare not taken into account.

The knowledge rules can only be applied ap-propriately by experienced microbiologists as-sisted by the use of literature and/or models pre-dicting the growth or inactivation of pathogens.

Simplified Identification Procedure forFood-Borne Microbial Hazards inFermented Sausage

Food Science and Technology Abstracts (ret-rospective 1969-1989 and current 1990-1999)were used as a database for the first levels ofhazard identification. The results are presentedin Table 7^4. With the rough hazard identifica-tion, only three pathogens (E. coll O157:H7,Salmonella, and S. aureus) involved in outbreakswere found. The ingredients used in this productare meat (pork, beef), pork backfat, salt, nitrate,and spices (pepper). With a detailed hazard iden-tification for the raw materials, all pathogenslisted in Table 7-2 are selected, even somepathogens that are indicated as not likely to bepresent (e.g., B. cereus). This is because the se-lection for Table 7-4 was based on literaturedata for the presence of these pathogens (even inlow numbers). Table 7-2 only presents informa-tion if there is a strong or a weak relationshipbetween the pathogen and the raw material.

Applying type 1 knowledge rules (survival ofpathogens in the end product) only results in theremoval of Campylobacter. This pathogen willnot survive fermentation processes due to thelow pH. Using type 2 rules removes B. cereus,Shigella, and Vibrio. These pathogens are oflittle concern in fermented sausages becausegrowth at low pH is not possible (B. cereus andVibrio spp.), or because the presence in the rawmaterials is only accidental (transmission viacontaminated persons).

Tabie 7-4 Results of the Hazard Identification Procedure Applied to Fermented Sausage

Knowledge Rules

Rough Hazard Identification Detailed Hazard Identification type 1 type 2 type 3

Aeromonas x xBacillus cereus xCampylobacter

Escherichia coll O157:H7 Escherichia coll O157:H7 x x xListeria monocytogenes x x x

Salmonella Salmonella x x xShigella x

Staphylococcus aureus Staphylococcus aureus x x x*Vibrio xYersinia x x

*Growth and toxin production of Staphylococcus aureus is usually restricted to the start of the fermentation process, when thereis a rather slow decrease in pH.

The pathogens likely to cause problems areAeromonas, E. coli O157:H7, L. monocyto-genes, Salmonella, S. aureus, and Yersinia.

To use type 3 rules, it was assumed that thepH of the final product was 5.0 ± 0. 1 and the Awwas 0.95 ± 0.01. If all types of knowledge rulesare applied, Aeromonas and Yersinia can be re-moved because usually high levels are neededfor food-borne infection, and growth in ferment-ing sausage is poor or not possible at all.61

The remaining pathogens are those from therough hazard identification (E. coli O157:H7, Sal-monella, and S. aureus) and L. monocytogenes.

Under the conditions described, growth andtoxin production by these pathogens are usuallyrestricted to the first 6-12 hours of fermentation.This can be followed by a gradual decline or sur-vival of the cells.25'26'31'33'43'47'56 Because of thepresence of competitive microorganisms,growth of S. aureus is seldom a problem.54 Thelow pH values will affect growth and survival ofL. monocytogenes .8,26,27,47 Jn contrast to the acid-adapted and acid-tolerant strains of Salmonellaand£. co//O157:H7.2'23'47

The knowledge obtained from the stepwiseidentification procedure provides a way to effi-ciently assess those hazards that need to be con-trolled during processing.

VALIDATION OF THE SAUSAGEFERMENTATION PROCESS FOR THECONTROL OF PATHOGENS

As a result of outbreaks of illness caused by S.aureus in fermented meat products,7 the Ameri-can Meat Institute formulated GMP for fer-mented dry and semi-dry sausages.3

The rapid decrease to low pH values, preferablylower than pH 4.0, that results in a sufficient inac-tivation of salmonellas and S. aureus turned out tobe of little effect against E. coli O157:H7,23'47 aswas demonstrated by the recent outbreaks of E.coli O157:H7 linked to the consumption of fer-mented sausages. The U.S. Department ofAgriculture's Food Safety and Inspection Service(USDA-FSIS) developed guidelines for validatinga 5-log reduction of E. coli O157:H7 in fermentedsausage.23 To meet the criteria set in the guideline,challenge tests should be performed to investigatethe behavior of this pathogen during the fermenta-tion process and a postpasteurization treatment.The information to perform a microbiologicalstorage test or a challenge trial can be gained fromthe literature or from laboratory test models (e.g.,the Food MicroModel). In most cases, the labora-tory models provide a more conservative viewthan the real product situation, as each parameter

has usually been studied under otherwise idealconditions.

Microbiological Challenge Tests

Laboratory simulation of what can happen toa food product during processing, distribution,and subsequent handling is an established tech-nique in the food industry. There are two mainreasons why this type of testing is carried out. Ingeneral, safety is of prime concern because thefood manufacturer must ensure that a productpresents a minimum risk to the consumer. Theother reason is to establish the shelf life or keep-ing quality of a product. In this case, quality is ofmost interest. In both cases, the product is nor-mally stored at specified temperatures for atleast the anticipated or prescribed shelf life. De-pending on the information available on rawmaterials, process, organisms of interest, and soforth, the product can be inoculated with (low)numbers of spoilage organisms and/or patho-gens before processing or storage. With theselaboratory simulation tests, which are manda-tory in some markets, and data from computermodels (e.g., Food MicroModel), it is possible todemonstrate that everything that can be reason-ably expected has been done.

Depending on the background for carrying out alaboratory simulation test (quality or safety), andwhether the product is likely to contain the typesof microorganisms of concern, these tests shouldbe named differently to avoid confusion. In a chal-lenge test, the product is inoculated with eitherpathogenic or spoilage organisms, depending onwhether consumer safety issues or product stabil-ity are of most concern. In a storage test, some-times called shelf life or keeping quality test, theproduct is not inoculated with microorganismsand the aim is to check safety and/or quality. Theexperimental plan and execution of a storage testmay, however, differ, depending on which of thefactors is to be studied.

Purpose of Microbiological Challenge andStorage Tests

One of the important tools used to ensure theproduction of safe food is the identification of

CCPs in an HACCP study. CCPs have to be con-trolled in order to eliminate hazards or minimizetheir likelihood of occurrence. It is essential thatthe identification and setting of criteria for CCPsare based on sound knowledge and relevant in-formation so that target levels and tolerance lim-its can be established. CCPs are under control ifthe criteria are met. Only if sufficient knowledgeconcerning the effect of intrinsic, extrinsic, andprocess factors on the safety of the product isavailable can the setting of criteria at each CCPbe done easily.

However, if information is scarce or unavail-able, data have to be obtained by storage testingand/or microbiological challenge testing. A mi-crobiological challenge test is an exercise tosimulate what can happen to a food product dur-ing processing, storage, distribution, and subse-quent handling, following inoculation with oneor more relevant pathogens.

It is important to realize when a microbiologi-cal challenge test should not be performed. If itis evident that the pathogen(s) will grow readilyin the product, a challenge test will be a waste oftime. Such pathogens must be eliminated byproper heat treatment or by preservation of theproduct.

Challenge testing will provide information onthe types of pathogens capable of growth in theproduct. With the data obtained, the risks of foodpoisoning can be assessed and the conditionsnecessary to prevent food poisoning can be de-termined. As a consequence, the safety of theproduct can be determined in terms of intrinsicfactors. Therefore, the information obtained isthe basis for setting criteria at relevant CCPs.

Challenge testing will provide informationabout the growth of selected pathogens in aproduct. However, the likelihood of presence ofthese pathogens in the product should be ob-tained from literature or databases.

Planning of a Storage Test and a ChallengeTest

The factors below, also relevant for the firstpart of an HACCP study, must be consideredwhen planning storage tests and challenge tests.

• microbiological status of the raw materials• product composition and limits for critical

hurdles• process design and parameters• packaging system• anticipated storage, distribution, and con-

sumer handling conditions• identification of the risks of food poisoning

from the product• existing knowledge of the types and prop-

erties of organisms relevant to the product

Carrying Out a Microbiological ChallengeTest

If the decision is made that a challenge test isrequired, and after deciding which organisms arerelevant for the test, a protocol has to be devel-oped. It is important to ensure that enoughsamples of the product are available at each timepoint of the investigation. The number ofsamples in the test depends on the heterogeneityof the product; all of the technical proceduresmust be reliable and reproducible.

There are two types of microbiological chal-lenge tests: process challenge tests and productchallenge tests. The objective of the process chal-lenge test is to determine whether or not a selectedpathogen survives a process (e.g., fermentation).The results from such tests are available directlyafter processing by enumeration of surviving mi-croorganisms. In some cases, such as ripening infermentation processes, analysis must be done af-ter a few weeks. Product challenge tests investi-gate whether or not the pathogens in artificiallycontaminated products can grow or under whatconditions unacceptable levels will be reached.

REFERENCES

1. Adams, M. R. & Moss, M. O. (2000). Food Microbiol-ogy, 2nd edn. Cambridge, England: Royal Society ofChemistry.

2. Adams, M. R. & Nicolaides L. (1997). Review of thesensitivity of different foodborne pathogens to fermen-tation. Food Control 8, 227-239.

3. American Meat Institute. (1982). Good Manufacturing

The test strains used (preferable a cocktail ofstrains) should be isolated from identical orsimilar products. During preparation of the in-oculum, stress conditions (that might reduce vi-ability) should be avoided. Because in practice,pathogens entering the product from the envi-ronment might be adapted to the intrinsic factorsof the product (e.g., poorly cleaned product en-vironment), it might be necessary to train themicroorganisms to growth at low pH or low Awin order to ensure reliable results.

Process Challenge Test for FermentedProducts Artificially Contaminated withE. coli O157:H7

Among the many different types of fermentedsausage, pepperoni is one of the most popularvarieties. In both the pepperoni and the Lebanonbologna process, E. coli O157:H7 survives thefermentation process.23'47

Performing challenge tests as described abovemight lead to the conclusion that other acid-toler-ant pathogens will also survive the fermentationprocess, which makes a postfermentation heattreatment necessary to produce a safe product.Another possibility, less suitable for raw meat butapplicable for items such as raw milk, is pasteur-ization of the raw materials. Zoning, in combina-tion with good manufacturing processes, shouldprevent postprocess contamination.

In the future, more research is needed to under-stand the behavior of (acid-adapted) pathogens infermentation processes. With this knowledge, hy-giene codes should be developed for all types offermentations in order to ensure safe products,both in industrial and in household technologies.

Practices: Fermented Dry and Semi-Dry Sausage.Washington, DC: American Meat Institute.

4. Antony, U., Moses, L. G. & Chandra, T. S. (1998). Inhi-bition of Salmonella typhimurium and Escherichia coliby fermented flour of finger millet (Eleusine coracand).World J Microbiol Biotechnol 14, 883-886.

5. Aryanta, R. W., Fleet, G. H. & Buckle, K. A. (1991).

The occurrence and the growth of microorganisms dur-ing the fermentation of fish sausage. Int J FoodMicrobiol 13, 143-156.

6. Aytac, S. A. & Ozbas, Z. Y. (1994). Survey on thegrowth and survival of Yersinia enterocolitica andAeromonas hydrophila in yogurt. Milchwissenschaft 49,322-325.

7. Bacus, J. N. (1986). Fermented meat and poultry prod-ucts. In Advances in Meat Research. Vol. 2, Meat andPoultry Microbiology, pp. 123—164. Edited by A. M.Pearson & T. R. Dutson. Westport, CT: AVI Publishing.

8. Beumer, R. R. (1997). Listeria monocytogenes: detec-tion and behaviour in food and in the environment. PhDdiss., Wageningen University Research Center, TheNetherlands.

9. Beumer, R., Bloomfield, S., Exner, M., Fara, G. M. &Scott, E. (1999). The need for a home hygiene policyand guidelines on home hygiene. Annali di Igiene 1 1 ,11-26.

10. Beumer, R. R., Vrouwenvelder, T. & Brinkman, E.(1994). Application of HACCP in airline catering. FoodControl 5, 205-209.

11. Bodnaruk, P. W., Williams, R. C. & Golden, D. A.(1998). Survival of Yersinia enterocolitica during fer-mentation and storage of yogurt. /. Food Sd 63, 535-537.

12. de Boer, E. & Hahne, M. (1990). Cross-contaminationwith Campylobacter jejuni and Salmonella spp. fromraw chicken products during food preparation. J FoodProt 53, 1067-1068.

13. Bonestroo, M. H., Kusters, B. J. M., de Wit, J. C. &Rombouts, F. M. (1993). The fate of spoilage and patho-genic bacteria in fermented sauce-based salads. FoodMicrobiol 10, 101-111.

14. Bryan, F. L. (1988). Risk of practices, procedures andprocesses that lead to outbreaks of foodborne diseases. JFood Prot 5 1,663-673.

15. Byaruhanga, Y. B., Bester, B. H. & Watson, T. G.(1999). Growth and survival of Bacillus cereus inmageu, a sour maize beverage. World J MicrobiolBiotechnol 15, 329-333.

16. Caplice, E. & Fitzgerald, G. F. (1999). Food fermenta-tions: role of microorganisms in food production andpreservation. Int J Food Microbiol 50, 131-149.

17. Centres for Disease Control and Prevention. (1993).Update: multi-state outbreak of Escherichia coliO157:H7 infections from hamburgers: Western UnitedStates, 1992-1993. MMWR 42, 258-263.

18. Chan, K. Y., Woo, M. L., Lam, L. Y. & French, G. L.(1989). Vibrio parahaemolyticus and other halophilicvibrios associated with seafood in Hong Kong. J ApplBacteriol 66, 57-64.

19. Corrales, M. T., Bainotti, A. E. & Simonetta, A. C.(1994). Survival of Vibrio cholerae Ol in common

foodstuffs during storage at different temperatures. LettAppl Microbiol 18, 277-280.

20. Cutter, C. N. (1999). Combination spray washes of sa-ponin with water or acetic acid to reduce aerobic andpathogenic bacteria on lean beef surfaces. J Food Prot62,280-283.

21. Dessie G., Abegaz, K. & Ashenafi , M. (1996). Fate ofSalmonella enteritidis and Salmonella typhimurium dur-ing the fermentation of siljo. East AfrMed Jl^, 432-434.

22. Dineen, S. S., Takeuchi, K., Soudah, J. E. & Boor, K. J.(1998). Persistence of Escherichia coli O157:H7 indairy fermentation systems. J Food Prot 61,1 602—1 608.

23. Ellajosyula, K. R., Doores, S., Mills, E. W., Wilson, R.A., Anantheswaran, R. C. & Knabel, S. J. (1998). De-struction of Escherichia coli O157:H7 and Salmonellatyphimurium in Lebanon bologna by interaction of fer-mentation pH, heating temperature, and time. J FoodProt6\, 152-157.

24. Encinas, J. P., Gonzales, C. J., Garcia-Lopez, M. L. &Otero, A. (1999). Numbers and species of motileaeromonads during the manufacture of naturally con-taminated Spanish fermented sausages (longaniza andchonzo). J Food Prot 62, 1045-1049.

25. Faith, N. G., Wierzba, R. K., Ihnot, A. M., Roering, A.M., Lorang, T. D., Kaspar, C. W. & Luchansky, J. B.(1998). Survival of Escherichia coli O157:H7 in full-and reduced-fat pepperoni after manufacture of sticks,storage of slices at 4° or 21 0C under air and vacuum,and baking of slices on frozen pizza at 135, 191 and 2460C. J Food Prot 61, 383-389.

26. Farber, J. M, Daley, E., Holley, R. & Usborne, W. R.(1993). Survival of Listeria monocytogenes during theproduction of uncooked German, American and Italian-style fermented sausages. Food Microbiol 10, 123-132.

27. Farber, J. M. & Peterkin, P. I. (1991). Listeriamonocytogenes, a food-borne pathogen. Microbiol Rev55,476-511.

28. Gahan, C. G. M., O'Driscoll, B. & Hill, C. (1996). Acidadaptation of Listeria monocytogenes can enhance sur-vival in acidic foods and during milk fermentation. ApplEnviron Microbiol 62, 3128-3132.

29. van Gerwen, S. J. C. & Zwietering, M. H. (1998).Growth and inactivation models to be used in quantita-tive risk assessments. J Food Prot 61, 1541-1549.

30. Griffiths, M. W. (1989). Listeria monocytogenes: its im-portance in the dairy industry. JSd FoodAgric 47, 1 33-157.

31. Hinkens, J. C., Faith, N. G., Lorang, T. D., Bailey, P.,Biige, D., Kaspar, C. W. & Luchansky, J. B. (1996).Validation of pepperoni processes for control of Es-cherichia coli O157:H7. J Food Prot 59, 1260-1266.

32. Hisa, K., Hayasi, K. & Sakaguchi, G. (1999). Proposalof HACCP systems for prevention of foodborne botu-lism. II. theoretical considerations in izushi productionto prevent Clostridium botulinum type E botulism basedon HACCP. JAntibact Antifung Ag - Jap 27, 3-1 1 .

33. Ihnot, A. M., Roering, A. M., Wierzba, R. K., Faith, N.G. & Luchansky, J. B (1998). Behaviour of Salmonellatyphimurium DT 104 during the manufacture and storageofpepperom. Int J Food Microbiol, 40 117-121.

34. Jacobs-Reitsma, W. F., Bolder, N. M. & Mulder, R. W.A. W. (1994). Cecal carriage of Campylobacter and Sal-monella in Dutch broiler flocks at slaughter: a one yearstudy. Poultry Sd 73, 1260-1266.

35. Johnson, E. A., Nelson, J. H. & Johnson, M. (1990).Microbiological safety of cheese made from heat-treated milk, part I: executive summary, introductionand history. JFoodProt 53, 441^52.

36. Johnson, E. A., Nelson, J. H. & Johnson, M. (1990). Mi-crobiological safety of cheese made from heat-treatedmilk, part II: microbiology. JFoodProt 53, 519-540.

37. Jyhshiun, L., In-Soo-Lin, Frey, J., Slonczewski, J. L. &Foster, J. W. (1995). Comparative analysis of extremeacid survival in Salmonella typhimurium, Shigellaflexneri, and Escherichia coll J Bacteriol 111, 4097-4104.

38. Kaspar, C. W. & Tamplin, M. L. (1993). Effects of tem-perature and salinity on the survival of Vibrio vulnificusin seawater and shellfish. Appl Environ Microbiol 59,2425-2429.

39. Kingamkono, R., Sjoegren, E., Svanberg, U. & Kaijser,B. (1996). Inhibition of different strains ofenteropathogens in a lactic-fermenting cereal gruel.World J Microbiol Biotechnol 11, 299-303.

40. Knorr, D. (1998). Technology aspects related to micro-organisms in functional foods. Trends Food Sd Technol9, 295-306.

41. Kunene, N. F., Hastings, J. W. & von Holy, A. (1999).Bacterial populations associated with a sorghum-basedfermented weaning cereal. Int J Food Microbiol 49, 75-83.

42. Leyer, G. J. & Johnson, E. A. (1992). Acid adaptationpromotes survival of Salmonella spp. in cheese. ApplEnviron Microbiol 58, 2075-2080.

43. Massa, S., Altieri, C., Quaranta, V. & DePace, R.(1997). Survival of Escherichia coli O157:H7 in yogurtduring preparation and storage at 4 0C. Lett ApplMicrobiol 24, 347-350.

44. Morioka, Y., Nohara, H., Araki, M., Suzuki, M. &Numata, M. (1996). Studies on the fermentation of softsalami sausage by starter culture. Animal Sd Technol67, 204-210.

45. Mortimore, S. & Wallace, C. (1998). HACCP: A Practi-cal Approach, 2nd edn. Gaithersburg, MD: Aspen Pub-lishers.

46. Motarjemi, Y. & Nout, M. J. R. (1996). Food fermenta-tion: a safety and nutritional assessment. Bull WHO 74,553-559.

47. Nissen, H. & Hoick, A. (1998). Survival of Escherichiacoli O157:H7, Listeria monocytogenes and Salmonellakentucky in Norwegian fermented, dry sausage. FoodMicrobiol 15, 273-279.

48. Northolt, M. D. (1983). Pathogenic micro-organisms infermented dairy products. Neth Milk Dairy J 37, 247-248.

49. Nout, M. J. R. (1994). Fermented foods and food safety.Foo d Res Int 27, 29 1-298.

50. Nout, M. J. R., Bakshi, D. & Sarkar, P. K. (1998). Mi-crobiological safety of kinema, a fermented soya beanfood. Food Control 9, 357-362.

51. Nout, M. J. R., Beernink, G. & Bonants-vanLaarhoven,T. M. G. (1987). Growth of Bacillus cereus in soya beantempeh. Int J Food Microbiol 4, 293-301 .

52. Nout, M. J. R. & Motarjemi, Y. (1997). Assessment offermentation as a household technology for improvingfood safety: a joint FAOAVHO workshop. Food Control8, 221-226.

53. Nout, M. J. R., Rombouts, F. M. & Havelaar, A. (1989).Effect of accelerated natural lactic fermentation of in-fant food ingredients on some pathogenic micro-organ-isms. Int J Food Microbiol 8, 351-361 .

54. Nychas, G. J. E. & Arkoudelos, J. S. (1990). Staphylo-cocci: then: role in fermented sausages. J Appl BacteriolSympSupp 19, 167S-188S.

55. O'Mahony, M., Mitchell, E., Gilbert, R. J., Hutchinson,D. N., Begg, N. T., Rodhouse, J. C. & Morris, J. E.(1990). An outbreak of foodborne botulism associatedwith contaminated hazelnut yoghurt. Epidemiol /«/104,389-395.

56. Ostergaard, A., Embarak, P. K. B., Wedell-Neergaard,C., Huss, H. H. & Gram, L. (1998). Characterization ofanti-listerial lactic acid bacteria isolated from Thai fer-mented fish products. Food Microbiol 15, 223-233.

57. Ozbas, Z. Y. & Aytac, S. A. (1996). Behaviour ofYersinia enterocolitica and Aeromonas hydrophila inskim milk during fermentation by various lactobacilli. ZLebensmittel Forsch 202, 324-328.

58. Poffe, R. & de Beeck, E. O. (1991). Enumeration ofAeromonas hydrophila from domestic wastewater treat-ment plants and surface water. J Appl Bacteriol 1 1 , 366-370.

59. Rhodehamel, E. J., Reddy, N. & Pierson, M. D. (1992).Botulism: the causative agent and its control in food.Food Control 3, 125-143.

60. Roberts, D. (1990). Sources of infection: food. Lancet336,859-861.

61. Roberts, T. A., Baird-Parker, A. C. & Tompkin, R. B.,eds. (1996). Microorganisms in Foods, 5, Microbiologi-cal Specifications of Food Pathogens. London: BlackieAcademic and Professional.

62. Sauer, C. J., Majkowski, J., Green, S. & Eckel, R. (1997).Foodborne illness outbreak associated with a semi-dryfermented sausage product. J Food Pr ot 60, 1612—1617.

63. Sharp, J. C. M., Reilly, W. J. & Coia, J. E. (1995). Es-cherichia coli 0 157 infection in Scotland: an epidemio-logical overview, 1984-1994. PHLS Microbiol Digest12, 134-140.

64. Shay, B. & Souness, R. (1995). Recent regulations im-pacting on the small goods industry. Food A ust 47,491-494.

65. Simango, C. & Rukure, G. (1992). Survival of bacterialenteric pathogens in traditional fermented foods. J ApplBacterial 73, 37^*0.

66. Smulders, F. J. M. & Greer, G. G. (1998). Integratingmicrobial decontamination with organic acids inHACCP programmes for muscle foods: prospects andcontroversies. Int J Food Microbiol 44, 149-169.

67. Swinbanks D. (1996). Outbreak of E. coli infection inJapan renews concern. Nature 282, 290.

68. Thomas C. & Beirne D. (1999). Microbiological safetyin processed vegetables. Int Food Hyg 10, 17-23.

69. Verotoxin-producing Escherichia coli 0 157. (1997).CDR Weekly 7, 409.

70. Warburton, D. W., Weiss, K. F., Purvis, U. & Hill, R. W.(1987). The microbiological quality of fermented sau-sage produced under good hygienic practices in Canada.Food Microbiol 4, 187-197.

71. West, P. A. (1989). The human pathogenic vibrios: apublic health update with environmental perspectives.Epidemiollnfm, 1-34.

72. Wu, Y. C., Kimura, B. & Fujii, T. (1999). Fate of se-lected food-borne pathogens during the fermentation ofsquid shiokara. J Food Hyg Soc Jpn 40, 206-210.

73. Yu, S. L. & Chou, C. C. (1987). Survival of some foodpoisoning bacteria in pickled vegetables stored at differ-ent temperatures. Food Sd China 14, 296-305.

CHAPTER 8

Microbiological Hazards andTheir Control: Viruses

Michael J. Carter and Martin R. Adams

INTRODUCTION

Viruses differ significantly from bacterialpathogens because they are obligate intracellularparasites and can replicate only within an appro-priate living host cell. The source of all viruses isthus a previously infected individual who shedsinfectious particles into his or her immediate en-vironment. Transmission to another host can bedirect, as in person-to-person spread (e.g.,through aerosols created by a sneeze or vomit-ing), or indirect, involving some other agent as acarrier for the virus. Food-borne transmission isof the indirect type. All viruses that infect via theenteric tract (and are shed in feces) are poten-tially capable of food-borne transmission. Directperson-to-person spread can also occur if the op-portunity arises; direct feces-to-mouth transmis-sion is common in children, and airborne spread,leading to contamination of distant surfaces, hasbeen reported.

Foods are contaminated with viruses as a re-sult of the distribution of fecal- or vomitus-de-rived viruses through the environment, eventu-ally contaminating the food or water of anotherpotential host. This process is highly variable. Itcould, for instance, result fairly directly from aninfected food handler contaminating food imme-diately before consumption, or it could be theend result of a prolonged distribution process

Acknowledgement — The authors would like to ac-knowledge Albert Bosch of the University ofBarcelona for the supply of unpublished information.

moving virus from fecal material to river and es-tuary and, perhaps eventually, to seafood. Gen-erally, these indirect routes require more timethan direct spread, and the stability of the virusparticle will be key to its successful distribution.Evolution has ensured that viruses of this typeare especially fitted for survival in transit, whichhas a direct influence on their ability to survivein processed food.

Water, either ingested directly or used as aningredient or washing agent during food pro-cessing, is the chief vehicle for disseminatingenteric viruses. It is also significant in carryingthese viruses to plants in the field (e.g., in irriga-tion water or sprays), and to the most significantvector of food-borne viral illness, molluscanshellfish. Most food-borne viruses survive wellin water. Their survival is assisted by high pro-tein content and high ionic strength, particularlycalcium and magnesium ions, which tend to sta-bilize the particles. These conditions are oftenfound in sewage and sewage-contaminated wa-ter. The adsorption of virus to suspended solidssuch as clays or organic matter can also protectthe particles from inactivation.

Viruses contaminating food are there as a re-sult of a dilution and distribution process thatmay have taken a significant time. Because vi-ruses have an absolute requirement for livingcells in which to replicate, they cannot increasein numbers during this distribution phase orwithin contaminated food during storage. Infact, the amount of contaminating viruses maydecrease during storage; this can be assisted by

treatment with heat, irradiation, chemicals, orother factors such as the pH changes associatedwith the fermentation of foodstuffs. Thus, thelevels of active viruses in foods must be ex-pected to be generally very low. Consequently,and with very few exceptions (mainly molluscanshellfish), viruses are only rarely detected infoodstuffs before consumption. In fact, it is onlymolluscan shellfish that are routinely monitoredfor the presence of viruses; in all other cases, acontaminating virus is usually inferred only ret-rospectively once consumption has led to ill-ness. Direct contamination of food from an in-fected food handler could deliver a moreconcentrated virus loading to a foodstuff, butbecause this could occur very sporadically, atany stage up to and including serving, testingfood for virus contamination from this source isgenerally not a viable option.

There are problems in assessing the impor-tance of viral food-borne illness. Figures gener-ally suggest that it is relatively low in adults,with the exception of the caliciviruses. How-ever, there is an almost complete lack of dataaddressing endemic levels of infection. Officialfigures refer almost entirely to "outbreaks" offood-borne virus illness, that is, to epidemic be-havior. Such outbreaks could result from the si-multaneous consumption of contaminated foodby many people, or equally from the infection ofa single person via contaminated food, followedby more direct person-to-person spread withinthe person's immediate surroundings. It is prob-ably significant that many outbreaks of infectionare observed in closed communities where hy-giene may not be advanced and where there isincreased potential for person-to-person spread,such as in childcare centers or nursing homes. Itis this subsequent transmission that renders theoriginal food-borne infection statistically "vis-ible." Food-borne infections such as these wouldnot be recorded if, for instance, they had oc-curred in a private home and affected one or onlya few relatively fit persons. Consequently, thefigures available probably represent only the tipof the iceberg as far as the incidence of food-borne virus transmission is concerned. Somemeasure of the extent of underreporting was

gained from the Infectious Intestinal DiseaseStudy conducted in England during the period1 993-1 996.48 This study concluded thatunderreporting was most severe for diarrhea-causing viruses, and that for every officially re-ported case of infection with Norwalk-like virus(NLV), there were actually 1,562 cases in thecommunity at large.

FOOD-BORNE VIRUSES

General Features

There are two broad classes of virus that mayinfect through the gut. The first class uses theenteric tract as the portal of entry to the body, butsubsequently spreads elsewhere around thebody. Examples of this type of infection wouldbe spread to muscle tissue of the skeleton orheart (as in the Coxsackie B viruses), to themeninges and central nervous system (polio andoccasionally other enteroviruses), or to the liver(hepatitis A and E viruses). Viruses in the sec-ond class are true inhabitants of the gut. Theyreplicate significantly only in this tissue and in-duce signs and symptoms of a typical "gastroen-teritis." This is mainly diarrhea, but can be ac-companied by differing degrees of vomiting.Viruses like this include the caliciviruses,rotaviruses, adenoviruses, and astroviruses. Fi-nally, there are a few agents that are detectedroutinely in the gut, such as the pico-birna vi-ruses. These are poorly characterized and do notseem to be associated with disease. The generalfeatures of these viruses, as well as their size,structure, and genome content, are provided inTable 8-1.

Difficulties in Culture and Diagnosis

The gut is a very specialized habitat in viro-logical terms; all viruses require living cells ashosts and are totally dependent on the processesthat their host cells are able to provide (e.g., forprotein synthesis and protein processing). Thereis a multiplicity of different cell types in the gut,each with a different role. This is reflected in adiffering complement of enzymes and surfaceproteins; cells may even vary in these properties

Virus

Caliciviridae

Reoviridae

Adenoviridae

Astroviridae

Picornaviridae

Members

Norwalk-like and Sapporo-like viruses

Rotavirus groups A-E;only groups A-C arefound in man. Group Cis rare and group B isuncommon except inChina.

Adenovirus group F, types40 and 41

Types 1-8

(Enterovirus)PoliovirusCoxsackievirusECHO virusEnterovirus

(Hepatovirus)Hepatitis A virus

Features

34 nm nonenvelopedparticles with distinctivemorphology (surfacecups).

Single stranded +vestranded RNA genome.

Multi-layered,nonenveloped 70 nmparticles, enclosingdouble-stranded RNAgenome. Only type A iscultivable and requires atrypsin supplement.

100 nm, nonenvelopedicosahedral particles, dsDNA genome. Cultivablein some cells (e.g.,Graham 293)

28 nm nonenveloped, starmotif may be visible.Single-stranded +ve RNA.Appearance can bevariable; may be mistakenfor calici or parvoviruses.Cultivable in differentiat-ing colonic cells withtrypsin supplement.

28 nm nonenvelopedparticles, single-stranded,+ve sense RNA. Growrelatively easily in avariety of human andprimate cell lines.

As above, cultivable withdifficulty.

Associated Illness

Norwalk-like: explosiveprojectile vomiting inolder children/youngadults. Noncultivable.

Sapporo-like: morecommon in youngerchildren believed to bemilder in effect,noncultivable.

Diarrhea common in theyoung, decreasing infrequency with age.Illness returns in theelderly.

Mild diarrhea, may beprolonged virus shed-ding. Mainly affectschildren.

Mostly infects children, buthigher serotypes arerarer and can causesignificant disease inadults. Relatively mild,but probably underesti-mated.

Mainly asymptomatic, caninduce muscle pains(Bornholm Disease),cardiomyopathy,meningitis, and CNSmotor paralysis (mostcommon with polio).

Hepatitis, mild in theyoung.

continues

Table 8-1 Food-Borne Viruses

Virus

Unclassified

Parvoviridae

Coronaviridae

Toroviridae

Members

Hepatitis E virus

Wollan, Ditchling. Cockleagents

Uncharacterized

Uncharacterized

Features

Calicivirus-like particlestructure, containingsingle stranded +ve RNAof unique genomicorganization.Noncultivable.

Smooth featureless 25 nmparticles. Single-strandedDNA genome. Poorlycharacterized,noncultivable.

Large enveloped virus,fragile, noncultivable.Food-borne transmissionunlikely in view of theirfragility.

Fragile, enveloped virus,may resemble Bernevirus in horses.Noncultivable, food-borne transmissionunlikely as above.

Associated Illness

Similar features to HAVbut more severe. Oftenfatal if contracted in latepregnancy.

Widespread shellfish-associated outbreaks,largely controlledthrough cooking.

Associated with neonatalnecrotising enterocolitis.

Unknown

Table 8-1 continued

at different stages during their differentiation.Furthermore, all gut cells are bathed in a solutionof the various products that are secreted by otherspecialized gut cells, notably, of course, pro-teases. These features make the gut a very diffi-cult cellular environment to mimic in culture.For instance, a virus might replicate in one par-ticular type of cell at one stage of its differentia-tion and could also require extracellular solubleproducts released from quite different cells en-tirely. For this reason, viruses that replicate inthe gut have been very difficult to culture in thelaboratory, and some still cannot be grown.Those that can be cultivated often require thatthe culture is supplemented with proteases, usu-ally trypsin,5 or even with duodenal juice.37 Spe-cialized cells are also sometimes needed (e.g.,differentiating colonic carcinoma cells forastro viruses).49 Viruses that penetrate beyond

the gut and invade other tissues are generally(but by no means always) simpler to cultivatethan those that reside in the gut.

This difficulty in cultivation has two impor-tant effects. First, virus detection can be prob-lematic. Usually, a cause of infectious gut illnesscan only be established in approximately half ofthe cases that are investigated. It is likely that alarge proportion of those cases of infection thatremain unidentified will in fact be found eventu-ally to be caused by viruses, but by viruses thathitherto have been missed because they are notculturable, and because there is no adequate di-agnostic test for this particular virus. The onlycatch-all method is electron microscopy, but thisdepends on a virus being readily recognizable byits shape and present in large numbers (manyworkers quote a minimum requirement of 106

particles per ml before observation becomes

even reasonably likely). Large viruses like therota- and adenoviruses are distinctive under themicroscope, but small, diffuse, or fuzzy virusesmay often be overlooked, and their identificationmakes high demands on the operator. Other testsmay be less demanding in use, but require char-acterized reagents such as antibodies for aggluti-nation or ELISA-based methods and primers forPCR-based detection. Thus, these tests can onlyfind viruses that are already characterized andknown; they cannot discover new ones. Theseshortcomings bias the relative detection fre-quencies of viruses associated with enteric in-fection. Although the extent of any such bias isnot clear, it is difficult to avoid the conclusionthat virus-associated enteric illness is likely to besubstantially underestimated, and that this prob-lem will be worse when considering the smaller,less distinct viral agents.

Second, if it is not possible to culture a virus invitro, then that virus can only be investigated invivo. If the only hosts are humans, such experi-ments become expensive and cumbersome, if notethically unacceptable. For this reason, many ofthe fundamental investigations into stability,transmission, and reinfection that are requiredhave not been done or have not been confirmed.

POTENTIALLY FOOD-BORNE VIRUSES

Table 8-1 summarizes the main viruses asso-ciated with enteric infection. Several of these areas yet uncharacterized and of necessity will notbe considered in this chapter. Any virus that isenterically transmitted is potentially transmis-sible via food, although person-to-person spreadmay be more common. This is especially true ofviruses that affect children and infants. Most en-teric viruses fall into this category. Infection isacquired early in life and becomes less commonthrough late childhood, adolescence, and adultlife, increasing in significance once more in theelderly. The significance of food-borne infectioncan thus depend on the types of food that areusually ingested by children. Chief among infec-tions of adults are the Norwalk-like viruses(NLVs, previously called the small round struc-tured viruses-SRSVs), and the Sapporo-like vi-

ruses (SLVs), two related members of a virusfamily called the Caliciviridae. Viruses of theNLV type are the most common viral cause ofexplosive outbreaks of projectile vomiting inadults. In recent years, reports of viral gastroen-teritis, particularly NLV outbreaks, have showna dramatic increase in England and Wales; 418cases were identified in 1992 and 2,387 in 1996.4

In 1994, reports of NLV outnumbered those ofSalmonella for the first time. This is probablydue in part to increased awareness and referralfor diagnosis, but it indicates the extent to whichsuch infections may have been missed in thepast, and probably continue to be missed today.

Increased surveillance for viral gastroenteritiswas introduced in the United Kingdom between1992-1995, and this has yielded some very in-teresting results. A preliminary data set was ob-tained for 2,149 of 2,680 outbreaks during thisperiod. Of these cases, 25% were attributed toSalmonella, 33% to NLV, 2% to rotaviruses, and0.5% each associated with astroviruses andsmall round (parvo virus) like agents. Twentypercent of cases were of unknown etiology, andapproximately half of these were believed due toviruses. The features of all of these viruses asso-ciated with food-borne transmission are de-scribed briefly in the following sections.

Caliciviruses (NLVs and SLVs)

NLVs and SLVs have different appearancesunder the electron microscope, and the relation-ship between them has been debated for sometime. SLVs have an obvious and unique struc-ture under the electron microscope; they appearto be covered in cup-like depressions, fromwhich the virus takes its name (calici = a cup).However, the NLVs appear fuzzy. Structuralstudies using cryo electron microscopy havenow revealed considerable underlying similarityin structure, and molecular studies have rein-forced these. It is now clear that these two agentsare related.11

NL Vs

NLVs were first identified following an out-break of enteric illness among adults in the town

of Norwalk, Ohio that has now given its name toall viruses of this type. NLVs are associated withsporadic outbreaks of diarrhea and vomitingamong young adults and older individuals. Thedisease is more common in winter than at othertimes of the year, and had been tentativelynamed winter vomiting disease in 1940, al-though its cause was not known at that time.NLV is characterized by an incubation period ofup to 48 hours, with illness lasting for 1-2 days.This virus differs from most other agents of viralgastroenteritis in that it mainly attacks adults andit frequently induces a high level of vomitingamong infected individuals. In fact, sudden, ex-plosive, projectile vomiting may be the first ob-vious symptom of infection. For this reason,many cases of NLV are actually identified atwork. Thus, the implications if a food handlershould be infected are obvious.

In addition, although there are multiple sero-types of NLV (probably more than eight), immu-nity to each seems to be short lived. Thus, indi-viduals may be protected for only a short period oftime (months) following an infection before theybecome susceptible to the same virus again.35

Some people appear to have an inherent resistanceto infection; these individuals remain symptomfree, even in the absence of antibody. However,sensitive persons may require several bouts of in-fection by the same virus before antibody levelsare sufficient for some protection. Even in middleage, only approximately 50% of people are serop-ositive. Although NLV does cause outbreaks ininstitutionalized settings where person-to-personspread can be significant, it is also associated withmultiple simultaneous exposure events where oth-erwise healthy adults consume virus-contami-nated food. Such outbreaks have minimal person-to-person transmission components, and 7-8% ofoutbreaks in the United Kingdom are described asmainly or chiefly food-borne in nature. However,this attribution is not particularly robust; the realfigure could be significantly higher.

NLV is spread chiefly by water. Significantoutbreaks of NLV have been associated with thecommercial manufacture of ice from contami-nated wells, soft fruits, and fruit juices. How-ever, chief among the sources of NLV outbreaks

is the consumption of raw or partially cookedshellfish. Molluscan shellfish are filter feeders,often farmed or growing naturally in estuarinelocations. In Europe, most estuaries are contami-nated with sewage effluent, and any viruses thatare shed in the water from communities livingupstream become concentrated within the bod-ies of the shellfish. Frequent outbreaks of gastro-enteritis have been associated with fecal bacteriathat was also concentrated in this way, but sim-ply relaying the shellfish in clean water (depura-tion) is often sufficient to allow them to purgethemselves of such bacteria. Unfortunately, thisprocess is inefficient in ensuring the removal ofviruses. Current European Union law regulatesthe sale of shellfish according to the levels ofcoliform bacteria they may contain, though thisis no guarantee of virological cleanliness. Con-sequently, each year, a number of outbreaks areattributed to the consumption of raw shellfish,particularly oysters. Outbreaks show a slightseasonality, for instance, shellfish seem to clearviruses less efficiently in the winter when seatemperatures are lower, and outbreaks may alsoincrease at such times. Eating habits can also bea contributory factor. For example, in the UnitedStates, there are often outbreaks around Febru-ary 14th and around Thanksgiving because oys-ters are traditionally eaten there at these times.

SLVs

SLVs cause illness more commonly in chil-dren and account for some 3% of hospital admis-sions for diarrhea in both the United Kingdomand the United States. Most children are serop-ositive by age 12 and seem to become infectedbetween three months and six years of age. Thedisease is particularly common in institutional-ized settings such as schools and daycare cen-ters. Incubation is between 24-48 hours, and ill-ness is usually mild and short lived, withdiarrhea tending to predominate. Any clinicaldifferences reported in the illness induced bySLV and NLV seem to reflect more the age ofthe patient than an intrinsic property of the virusitself. On those occasions where SLV may causeillness in adults, the signs and symptoms are in-distinguishable from those of NLV.

An assessment of the stability of NLV is par-ticularly difficult because the virus can only becultivated in volunteers. However, despite thislimitation, a few valuable studies have addressedthis point, and it is worth commenting on thesensitivity of NLV to chlorine used in the treat-ment of potable water. Reports from the UnitedStates show that the virus resists treatment at apeak level of 5 mg/1 for 30 minutes. Thus, NLVcan survive water chlorination to this level andhas even been detected in tap water in the UnitedKingdom (although it is not known if it was in-fectious at that point). NLV is inactivated at lev-els of 10 mg/1, and this is used to decontaminatewater supplies if contamination is suspected.The virus is relatively more chlorine resistantthan polio and human rotaviruses (rotavirus isinactivated at levels greater than 3.5 mg/1).26

NLV is acid stable, surviving at pH values aslow as 2.7 for three hours at room temperature,and also relatively heat stable (resists 60 0C for30 minutes).13 Because the virus cannot begrown, feline calicivirus (FCV) is frequentlyused as a model.44 Although it is true that the fe-line virus will probably mimic the behavior ofNLV in terms of where it tends to accumulate, itis a relatively poor model for stability. For ex-ample, FCV is considerably more acid labilethan NLV. Like all other enterically transmittedviruses, NLV lacks an envelope and thus tendsto resist lipid solvents that would otherwise inac-tivate viruses by stripping their lipid envelopes.For instance, NLV is not inactivated by 20%ether at 4 0C for 18 hours.13

Rotaviruses

Rotaviruses are members of the familyReoviridae (Table 8-1) and account for some3.5 million cases of diarrhea in the United Stateseach year. This equates to 35% of hospital ad-missions for diarrhea each year. Even in a devel-oped country like the United States, approxi-mately 120 children die each year from thisvirus, and fatalities are probably far more nu-merous in less developed countries. The peakage for illness is between six months and twoyears; by four years of age, most people have

been infected.27 Immunity to rotaviruses is longlasting and secretory immunoglobulin A (IgA)in the gut plays an important role.

Group A rotaviruses exist in nine serotypes,and variations in other proteins increase the anti-genie diversity of this group. However, as timeprogresses, individuals are exposed to many dif-ferent variants of rotavirus, and immunity gradu-ally accumulates. Thus, frequency of illness de-creases with age. However, silent secondaryre-infections can occur throughout life (as inparents caring for infants), thereby providing an-other means for the virus to spread in the com-munity. Illness usually develops after an incuba-tion period of four to seven days and presents asdiarrhea and vomiting lasting approximately oneweek. In the United Kingdom, only some 2% ofoutbreaks were attributed as food-borne. The vi-ruses are stabilized by calcium ions, which pro-mote integrity of the particle.42 Ionic detergentscan reduce infectivity (e.g., 0.1% sodiumdodecyl sulphate, or SDS), but non-ionic deter-gents can actually increase virus titre by break-ing up aggregates and dispersing the particles.Viruses are very resistant to acid conditions andare stable over a wide pH range from 3.0 to 9.O.14

They can retain infectivity for months at 20 0C42

and also resist heating to 50 0C. Rotaviruses tendto survive desiccation well and are reduced in ti-tre by less than 10-fold during drying (A. Bosch,personal communication). Once dried, the virussurvives well on either porous or nonporous sur-faces, being relatively unaffected by temperatureor relative humidity. Infectivity of dried virusesreduces in titre by just more than 2 logs over a 60-day period, regardless of substrate, temperature,relative humidity, or the presence of accompany-ing fecal material. This makes rotavirus one of themost stable of enteric viruses, second only to hepa-titis A virus in this respect.1

Adenoviruses

There are 47 serotypes of adenoviruses, butonly types 40 and 41 cause enteric illness.Adenoviruses account for some 5-20% of U.S.hospital admissions for diarrhea, mainly in chil-dren below two years of age.27 Incubation lasts

three to ten days, and illness (watery diarrhea)may last for one week. However, prolongedshedding of the virus has been reported. As chil-dren age, exposure to adenovirus infectiongradually increases the levels of immunity. Only20% of children below six months of age haveantibody to these viruses, but by age three, thishas risen to 50%.27 In the United Kingdom, ad-enovirus-related outbreaks are extremely rare;only one case was identified between 1992-1995, and this was not believed to be food-borne.4 Clearly, this is in stark contrast to theantibody prevalence figures and implies that in-fection may be mainly in the community anddoesn't follow outbreak behavior.

There are few studies that have specifically ex-amined the stability properties of the entericadenoviruses, but most workers agree that the vi-ruses are certainly stable between pH 5 and 9 andup to 45 0C.17 Infectivity is rapidly (10 minutes)lost above 56 0C, and this is associated with par-ticle disintegration.40 Other workers have exam-ined the stability of the enteric adenovirus type 40,when dried on to fomites.1 Adenovirus is one ofthe least stable of the enteric viruses under theseconditions, and its behavior seems to resemble thatof poliovirus rather than rota-, astro-, or hepatitisA viruses, which tend to be more robust. Desicca-tion itself causes a 100-1000-fold decrease in titre.However, infectivity persists for seven days onnonporous substrates, it is stabilized by the pres-ence of accompanying fecal material, and it is un-affected by levels of relative humidity. Adeno-viruses survive less well on porous surfaces, andthe accompanying material seems to decrease itsstability. Survival is better at 4 0C than at 20 0C oneither surface. In liquids, adenovirus stabilityclosely resembles that of the astroviruses, whichare small, round RNA containing viruses respon-sible for enteric infections mainly in the young ofanimals and humans. Adenoviruses survive rela-tively well in tap water; log titre reduction valuesof 3.2 were observed after 60 days at 20 0C. Thevirus is also disinfected by free chlorine, givinglog titre reduction values of 2.5 and 3 after twohours in the presence of 0.5 and 1 mg/1 free chlo-rine, respectively. Again, this behavior mimicsthat of human astroviruses.2'3

Astroviruses

Astroviruses (serotypes 1-8) account forsome 5% of hospital admissions in the UnitedStates, almost entirely of children.27 By sevenyears of age, 50% of children are already serop-ositive for the most common serotype (type 1),and this reaches 75% by age 10. Illness is gener-ally mild and lasts some two to three days afteran incubation period of similar length. This hasled many researchers to dismiss these viruses ascausative agents of significant disease in hu-mans. However, this is not necessarily true, firstbecause the higher serotypes are less common,thus infection is delayed and occurs among olderchildren and adults. Preexisting antibody toother serotypes may modify the severity of anyresulting illness but may not prevent the occur-rence of clinical disease. In Japan in 1995, 1,500older children and their teachers were affected ina widespread food-borne outbreak caused byastrovirus type 4,33 and symptomatic illness hasalso been seen in adults in the United Kingdomand France. Second, astrovirus diagnosis canpresent particular problems that may lead to anunderestimation of the number of cases. Diagno-sis still largely relies on electron microscopy,but particle morphology alone is not necessarilya good guide. Astroviruses may frequently bemistaken for small round (parvovirus-like)agents and even for NLVs.23 This is a significantfinding, because a recent survey in the UnitedKingdom identified only nine outbreaks associ-ated with astrovirus (and a further two mixed in-fections that also contained NLV), of which ap-proximately 20% were thought to befood-borne.4 A further four outbreaks of illnesswere associated with small round virus, an agentusually assumed to be parvovirus and routinelydismissed as the causative agent of the outbreak.However, half of these cases were believed to befood-borne, and in view of the potential formisidentification, these agents require furtherinvestigation. Should these cases actually repre-sent misidentified astroviruses, then the propor-tion of astrovirus outbreaks classified as food-borne could be as high as 30-45%. Recently, anELISA-based detection kit has been produced,

which could help to answer these questions if itis widely adopted.

Astroviruses survive desiccation well, drop-ping by only 10-fold with or without accompa-nying organic material. If the virus is dried on toa nonporous surface, accompanying fecal mattercan boost relative survival by 10-100-fold, withinfectivity persisting for up to 65 days and prob-ably much longer at 4 0C. Astro virus survival isgreatly reduced at increased temperatures andcan decay completely within 10 days at 20 0C ona nonporous surface. Adenoviruses surviveequally well on porous surfaces at 4 0C regard-less of the presence or absence of fecal material,infectivity persisting for more than 90 days. Thissuggests that environmental temperature mayplay an important role in the seasonality ofastrovirus outbreaks because indirect transmis-sion would be assisted by decreased temperature(A. Bosch, personal communication).

Astrovirus survives well in dechlorinated(tap) water3 and reduces in titre by 100-fold after60 days at 4 0C. This is increased to a 3.2 logreduction after 60 days at 20 0C, which is a verysimilar stability to that observed for humanrotavirus and adenovirus type 40, where thesame reduction was seen at 20 0C. Free chlorineseems more effective at disinfecting astrovirus-contaminated water than water contaminatedwith hepatitis A or human rotavirus. Astrovirustitres were reduced by 2.5 logs after one hour at afree chlorine concentration of 0.5 mg/1. This in-creased to 3 logs in the presence of 1 mg/1 freechlorine. However, in both cases, residual infec-tivity was still detected after two hours of treat-ment when the log titre reduction was 4.17.3

Hepatitis Viruses

Hepatitis viruses A and E are classified differ-ently. Hepatitis A virus (HAV) is the only mem-ber of a single virus genus (Hepatovirus) of thePicornaviridae. Hepatitis E virus (HEV), how-ever, has some structural features resembling thecaliciviruses, but its unique genomic organiza-tion means that it cannot be classified easilywithin any of the existing virus families.

Both viruses spread predominantly throughwater and are concentrated by molluscan shell-fish. HAV is not uncommon, but HEV is rare indeveloped countries. It does, however, occur inepidemic form in India and in the former SovietUnion. The worst outbreak of HEV involved30,000 people in New Delhi in 1955. More lim-ited shellfish-associated outbreaks occur spo-radically around the Mediterranean. In theUnited Kingdom, HEV is limited to returningtravelers. Clinical features of both viruses aresimilar, although HEV tends to be more severeand can be fatal in pregnancy. Convalescencemay be prolonged (8-10 weeks), and some 15%of cases of HAV may follow a relapsing courseover 12 months or more.

The spread of picornaviruses has been af-fected profoundly by human activities. In formertimes, when the quality of water could not beguaranteed, infection occurred early in lifethrough exposure to virus-contaminated water.Under these conditions, infections tended to bemild (often subclinical) and were endemic in thesociety. However, where water purification andother public health measures have been imple-mented, the possibilities of infection for all ofthese viruses are reduced, which has had the ef-fect of increasing the mean interval betweencontacts with the virus. As a result, infection isdelayed, and thus occurs predominantly in olderindividuals.

This infection delay is well demonstrated incountries where sanitation has improved overrecent years.19'28 In Hong Kong in 1979, 30% ofthose people under 30 years of age had been in-fected with HAV; by 1989, this number was 9%and is still falling. Presumably, this will progresseventually to resemble the situation in the devel-oped countries where the vast majority of per-sons older than 30 have no antibody to HAV(i.e., they have never been infected).18 This issignificant for the course of the infection be-cause the virus is more severe if it is contractedin adulthood. People older than 30 years of ageaccount for only some 30% of the cases of HAV,but nearly 80% of the deaths occur in this agerange.19 Shellfish will, of course, concentrateany viruses shed in sewage and contaminating

their natural habitat, including picornaviruses,and perhaps more significantly, HAVs andHEVs. Significant outbreaks of illness havebeen noted in Singapore and have been attrib-uted directly to virus contamination of shell-fish;47 similar outbreaks that are mainly attribut-able to direct virus contamination of water occurwith HEV.28 Consequently, as water quality im-proves, it is expected that persons growing olderin the absence of HAV infection would developa taste for shellfish and shellfish products (in-cluding fermented foods) and become exposedto the virus later in their lives. They would haveno residual immunity from childhood infections(see astroviruses above), which could pose a riskof severe disease later in life.

HAV is more stable to acid than other gut vi-ruses of the picornavirus family. It retained vir-tually all activity following 120 minutes at pH1.0 and was still viable (although reduced in ti-tre) after five to eight hours.41 Other enterovi-ruses were inactivated virtually completelywithin two hours at this pH. HAV is also rela-tively more heat stable and survives heating at60 0C for 60 minutes,10'15'21 being only partiallyinactivated after 12 hours at this temperature atneutral pH.36'43 In the absence of divalent cat-ions, 50% of particles will disintegrate within 10minutes at 61 0C, although in the presence of 1M MgCl, this is not achieved until 81 0C. Theequivalent temperatures for poliovirus are 43 0Cand 63 0C, respectively, and polio appears farmore sensitive to desiccation and storage in thedry state than HAV. This suggests that poliovi-rus is an inadequate model for the survival ofenteric viruses in general, and especially forHAV.1'46 HAV does, however, become rapidlyinactivated (in minutes) at temperatures higherthan 98 0C and can also be destroyed by chlorineand hypochlorite (10-15 ppm residual chlorineafter 30 minutes; 3-10 ml/1 hypochlorite 20 0Cfor 5-15 minutes). However, HAV is more chlo-rine resistant than other picornaviruses.2'38 Ingeneral, HAV is the most stable of theenterically transmitted viruses. It survives desic-cation with minimal decrease in titre (0.5 log)and, once dried, shows virtually no reduction intitre during a seven-day period regardless of the

surface onto which it has been dried, tempera-ture or relative humidity of storage, or the pres-ence of accompanying fecal material. Titres arereduced by only 1-2 logs during the next 60 daysafter drying.1 These data suggest that the sur-vival of HAV is likely in fermented foods, al-though it has not been rigorously assessed, andexposure to the virus from the consumption ofshellfish or fermented foods prepared from themremains a possibility.

ASSESSMENT OF VIRUS RISK

Contamination of Foods

Virus levels will not increase during the fer-mentation of food; therefore, virus content ofthese materials becomes an issue only wherethere is significant contamination of the food be-fore it is fermented. Most food-borne viruses areresistant to both acid conditions and mild heattreatment and could survive these aspects offer-mentation processes. NLVs represent the great-est risk of gastroenteritis for adults because theyare not likely to be immune. Stability data forthese viruses are scarce because of the difficultyin assessing their survival, but it seems unlikelythat they would be more resistant to physicaltreatments than HAV. Thus, in assessing the po-tential for virus survival in food, it would seemthat procedures that would inactivate HAVprobably give the greatest protection against vi-rus dissemination in the food. The other virusesthat could be present in food are largely infec-tions of childhood. With the exception of fer-mented cereals and dairy curds in some cultures,children may not eat much fermented food andthus the acquisition of infection from this source(even if the viruses survive) is likely to be rare.Adults consuming food that has been contami-nated with these viruses would be protected byresidual immunity resulting from childhood in-fection. HAV poses a serious threat to adultswith no preexisting immunity. Because this isthe virus that is hardest to remove (and thus mostlikely to survive), HAV contamination of fer-mented foods (especially of shellfish-derivedfoods) could pose an increasing threat in areas

where sanitation is rapidly improving and serop-ositivity to HAV is falling in adults.

Improved knowledge of the types of food thatare regularly contaminated with viruses shouldpermit an assessment of the relative risks associ-ated with particular foods. In the United King-dom, rigorous investigation often fails to iden-tify the food vehicle of transmission. However,accepting the significance of NLVs as the mostcommonly identified cause of illness in adults, arecent survey found that 35 (65%) of the NLVoutbreaks could be attributed to a particularfoodstuff, the most frequent being oysters.4 Inthose cases where a particular food was impli-cated, virus was detected in the food itself inonly two cases, both oysters. This percentage re-flects the generally low level of virus contamina-tion in other types of food. Other food vehiclesimplicated were diverse, including sandwiches,pies, fresh fruit, fruit and vegetable salads,gateaux, fish, lobsters, and prawns. Many of thedishes were not served alone, and the commonfeature could be fresh fruit or salad vegetablesthat were often served as a garnish if not a maincomponent of the meals. It is not possible to as-sess how many of these outbreaks were associ-ated with contamination by food handlers imme-diately before serving, and how many mighthave resulted from virus contamination of thefood before its preparation. Certainly, food han-dlers were suspected in 19 cases, but was con-firmed in only four. Viral illness has been asso-ciated with intrinsic contamination of fruit in thepast,6'31 and the worldwide trade in fruit and veg-etables may well increase such a risk in the fu-ture. It is generally held, in the United Kingdomat least, that there are two major patterns oftransmission of food-borne viruses — those dueto intrinsic contamination of shellfish with virus,and those due to other foods probably involvingcontamination by food handlers. It is not generallybelieved that virus contamination of vegetablesand fruit at source is a major contributor to food-borne viral illness in the United Kingdom.

Food-borne spread of HAV is not common. Inthe United Kingdom, during the 1980s, sevenoutbreaks were attributed to shellfish and a simi-lar number to food contamination by handlers.

More recently, out of 19,000 cases (mostly in-volving children), only 0.5% were associatedwith food, with shellfish again being the majorculprit. This risk is far more significant else-where where substantial outbreaks have been at-tributed to shellfish consumption. An outbreakassociated with clams in Shanghai may havebeen the largest recorded outbreak of food-borneillness ever.20 HEV occurs in the United King-dom only in one or two cases per year, entirelyassociated with travelers returning from areas inwhich the disease is endemic. Waterbornespread would seem to be the major vehicle oftransmission, although shellfish-associatedcases have been reported.

Virus Survival in Fermented Foods

There are few data that are directly concernedwith the survival of viruses in fermented foods.Studies conducted in fermented meats haveshown that the viruses responsible for foot andmouth disease (i.e., FMDV), African swine fe-ver, and hog cholera do not survive the produc-tion process.29'34 However, these viruses are ofveterinary importance and are known to be morelabile than human enteric viruses. FMDV is acidlabile, is inactivated below pH 6, and is not likegut viruses of the same family such as polio,whereas the viruses causing hog cholera and Af-rican swine fever are both enveloped and there-fore more susceptible to the adverse conditionsencountered during fermentation.16 When modelenteric viruses such as polio, Coxsackie, andechovirus, which are more relevant to humanhealth, were used, they were found to persist inhigh titres virtually unaffected by the processingthat reduced the pH to 5. 0-5. 4. 12'22'25 Studieswith bacteriophage and simian rotavirus in fer-mented cereals that are more weakly bufferedand consequently have a lower pH have alsodemonstrated the ability of viruses to survivewell in fermented products.32'50 This relativepaucity of information does, however, point tothe difficulty of controlling risk from food-borneviruses through fermentation, though furtherwork is clearly needed. Based on all of the infor-mation available, however, some general infer-

ences can be drawn concerning the risks posedby viruses and their control.

First, the type of food on which the product isbased will be an important factor. Shellfish are anobvious high-risk material because they often livein sewage-contaminated estuarine waters and con-centrate available viruses in their tissues. The ini-tial preparation of shellfish involves shucking theshellfish flesh, and even if this is then washed, it isunlikely to remove virus contamination from theshellfish gut. Fermented foods prepared from or-ganisms that consume shellfish (such as octopusor scavenging crabs) may be contaminated indi-rectly from this source. However, viruses are notbelieved to be retained with the same efficiency inthese creatures, so the initial risk is presumablymore short lived.

Vegetable produce may be contaminated ei-ther on the surface (e.g., by handling, washing,or spraying with contaminated water)8 or moredeeply within the tissues (e.g., resulting from theuptake of viruses contaminating irrigation wa-ters used in cultivation). This is more likely inthe case of produce with a high water content,such as celery, pumpkins, cucumbers, and othersoft fruits. Surface contamination can be effec-tively removed by peeling provided the peelingsare not simply allowed to elute any virus into thewashing water and thus redistribute the materialacross freshly peeled surfaces. Similarly, wash-ing itself can be very effective, and efficiencycan be increased if small quantities of (prefer-ably ionic) detergent can be used.

Where produce is contaminated with fecalmaterial, viruses may be dried on to the surfaceas aggregates with other organic material. Thistends to promote survival on nonporous surfacessuch as a waxy plant cuticle. In these cases, theuse of detergent could not only remove the ag-gregate, but also disperse it, thus increasing thenumber of potential infectious units present. Un-der these conditions, thorough rinsing becomesessential. Potable water washing alone can re-duce surface contamination by bacteria by 10-100-fold, and the use of hypochlorite is evenmore effective. Common commercial washingprocedures in the developed world use 100 ppmhypochlorite (yielding 30-40 ppm free chlorine)

at pH 6.8-7.1 and 4 0C and a contact time of twominutes. Other materials such as soft fruits thatcould be damaged by this process can be sprayedor immersed for only 10 seconds in 15-20 ppmfree chlorine.7 Tests have found that in general,hypochlorite under these conditions is very ef-fective against viruses in suspension;24 however,its efficacy against viruses that are adsorbed to asurface or in the presence of other organic mate-rials cannot be assumed as great. Studies suggestthat hypochlorite may not be as effective as as-sumed when it is used to control HAV that wasartificially introduced on to strawberries. Hy-pochlorite treatment was approximately 10-foldless effective against this virus on the fruit thanthe same virus in suspension, and full controlcould not be achieved below levels that wouldrender the fruit inedible.45 Furthermore, thismethod of cleansing must be limited to an effecton surface contamination only; it could clearlynot affect any deeper contamination within theplant tissues.

Fermentation frequently involves salting, par-ticularly fermented shellfish products producedin southeast Asia, where salt levels of up to 30%by weight can be used, producing a saturatedbrine.39 Salt (sodium chloride) itself is generallynot injurious to virus particles lacking an enve-lope; other ions are more problematic, with ce-sium having an adverse and sometimes irrevers-ible effect on virus polymerase function. Highsalt (especially under acid conditions) can leadto the precipitation of proteins; this is likely tooccur and will be aided by a generally high pro-tein concentration in the environment. Precipita-tion on this scale would certainly cause the pre-cipitation of virus particles and a considerablereduction in their infectivity. Precipitated vi-ruses can redissolve should the salt concentra-tion be reduced at a later stage, but residualclumping and loss of infectivity would mean thatthe original titre will not be recovered. Othersalts have a stabilizing effect on virus particles,particularly divalent cations; magnesium stabi-lizes polioviruses and calcium stabilizesrotaviruses.

pH reduction in fermentation is typical, andpH could fall to as low as 3.8. However, all en-

teric viruses are able to infect through the gut,and thus all must be designed for passagethrough the acid conditions of the stomach. In-deed, stability to acid (pH 4.0) remains one ofthe routine tests for enteroviruses, differentiat-ing them from the morphologically identical rhi-noviruses that infect via the nasal mucosa andare not required to possess acid stability. Acidresistance is thus a common feature of all of theviruses in these groupings, and acid productionper se is not likely to reduce virus titre signifi-cantly unless accompanied by heat. This is sup-ported by the limited number of studies done onvirus survival in acid-fermented foods. Nout etal.32 used the bacteriophage (MS-2) as a modelfor human viruses and found that it survivedmuch better than bacteria in a fermented por-ridge at pH 3.8; its numbers declining by ap-proximately 0.1 log cycle per hour. Even bettersurvival was found in work using the simianrotavirus SA 1 1 , a good model for the behaviorof unculturable human rotavirus. In 24 hours at30 0C, the virus titre decreased by 0.25 logcycles at pH 4.0 and by 1 log cycle at pH 3.3.50

Significantly, this work found the effect to bepurely pH-related, with no difference when dif-ferent acidulants were used. The enhanced anti-microbial effect of weak organic acids seenagainst bacteria clearly does not operate withnonenveloped, nonmetabolically active virusparticles.

During fermentation, autolysis of the tissueswill lead to protease release. Gut viruses are re-quired to possess some degree of protease resis-tance because infection through the gut must ex-pose them to these enzymes. In fact, some seemto have evolved to require the addition of pro-teases such as trypsin during their culture invitro. Trypsin, however, is a specific enzymerecognizing certain sites in the protein, and allviruses are susceptible to prolonged proteasetreatment. This is especially true of broad-spec-trum proteolytic enzymes such as protease K andpronase, and plant enzymes such as bromelainand papain. The enzymes released during au-tolysis are also broadly reactive against proteins.Regardless of the context of the peptide bond,they too are able to destroy virus particles even-

tually. Detergents and saponifiers will also assistthis action if present. The addition of protease-containing fruit or fruit juice (especially pine-apple) would also aid this effect in fermentedshellfish products such as Plaa-mam and Khem-mak-nat.39 Virus particles would, however, beprotected from this process if they are present asaggregates or inclusions (perhaps within shell-fish gut tissue) or surrounded by a high localconcentration of protein.

Finally, there is heat treatment. The ability ofa heat treatment to eliminate risk from a patho-genic organism will depend not only on the in-trinsic heat sensitivity of the organism, but alsoon the initial numbers of the organism present,the heating menstruum, the temperature, and thetime of exposure. Although there is a wealth ofinformation on how these factors can affect bac-terial survival, the thermal inactivation kineticsof viruses have not been subject to the same kindof scrutiny. A temperature of 65 0C is ineffectiveagainst most of the enteric viruses; certainlyastroviruses, HAV, and rotaviruses can survivesuch treatment even if adeno viruses and polio vi-ruses may be substantially reduced in titre. A re-cent study9 has shown that less than 0.5 minutesat 85 0C produced a reduction of more than 5 logcycles in the titre of HAV in three different dairyproducts, whereas heating at conventional pas-teurization temperatures of 71-73 0C required13-18 minutes to achieve the same effect. Someprotective effect from fat was seen in cream con-taining 18% fat, but no significant differenceswere discernible between skimmed and whole(3.5% fat) milk.

The failure to see outbreaks of viral illness as-sociated with correctly pasteurized milk indi-cates the very low association of human viruseswith this product. Thus, by extension, fermentedmilk products are likely to be relatively safe withregard to viral infections, particularly when op-erations such as milking and fermentation aremechanized, thus reducing the risk of contami-nation from food handlers. FCV could not becultured from cockles immersed in boiling waterfor one minute or longer, during which time theaverage internal temperature reached 78 0C. Im-mersion for 30 seconds, which achieved an aver-

age internal temperature of approximately 63 0C,produced a 2-log reduction in titre.44 HAV (andpolio virus) were completely eliminated fromcockles that achieved an average internal tem-perature of 85-90 0C for one minute. These datawere used as the basis for setting heat-treatmentregulations for cockles in the United Kingdom(90 0C for 90 seconds).30 In some products,though, the temperature treatment necessary toeliminate risk may be incompatible with productquality, although there is some evidence that in

REFERENCES

1. Abad, F. X., Pinto, R. M. & Bosch, A. (1994). Survivalof enteric viruses on environmental fomites. ApplEnviron Microbiol 60, 3704-3710.

2. Abad, F. X., Pinto, R. M. & Bosch, A. (1997). Disinfec-tion of human enteric viruses on fomites. FEMSMicrobiol Lett 1 56, 1 07-1 1 1 .

3. Abad, F. X., Pinto, R. M., Villena, R., Gajardo, R. &Bosch, A. (1997). Astrovirus survival in drinking water.Appl Environ Microbiol 63, 3119-3 122.

4. Advisory Committee on Microbiological Safety ofFood. (1998). Report on Foodborne Viral Infections,London: The Stationery Office.

5. Albert, M. J. & Bishop, R. F. (1984). Cultivation of hu-man rotaviruses in cell culture. JMed Virol 13, 377-383.

6. Anon (1997). Hepatitis A associated with consumptionof frozen strawberries. WHO Surveillance Programmefor Control of Foodborne Infections and Intoxicationsin Europe 54, 5.

7. Beuchat, L. R. (1992). Surface disinfection of raw pro-duce. Dairy, Food Environ San 12, 6-9.

8. Beuchat, L. R. (1996). Pathogenic microorganisms as-sociated with fresh produce. JFoodProt. 59, 204-216.

9. Bidawid, S., Farber, J. M., Sattar, S. A. & Hayward, S.(2000). Heat inactivation of hepatitis A in dairy foods. JFood Prot 63, 522-528.

10. Bradley, D. W., Fields, H. A., McCaustland, K. A.,Cook, E. H., Gravelle, C. R. & Maynard, J. F. (1978).Biochemical and biophysical characterisation of lightand heavy density hepatitis A virus particles: evidencethat HAV is an RNA virus. JMed Virol 2, 175-187.

11. Carter, M. J. & Cubitt, W. D. (1995). Norwalk and re-lated viruses. Curr Opinion Infect Dis 8, 403-409.

12. Dethmers, A. E., Rock, H., Fazio, T. & Johnston, R. W.(1975). Effect of added sodium nitrate on sensory qual-ity and nitrosamine formation in thuringer sausage. JFood Sd 40, 491-495.

acid (fermented) foods, the heat resistance of vi-ruses is much reduced.50

Viruses clearly pose some special problems.The available data suggest that a fermentationprocess alone will not ensure safety from viralinfection, although the highest risk is associatedonly with a very limited number of substrates.Further studies are clearly indicated to explorethe integrated effect of all aspects of processingon virus survival in order to establish a realisticestimate of risk.

13. Dolin, R., Blacklow, N. R., DuPont, H., Bushco, R. F.,Wyatt, R. G., Kasel, J. A., Hornick, R. & Chanock, R.M. (1972). Biological properties of the Norwalk agentof acute non-bacterial gastroenteritis. Proc Soc Exp BiolMed 140, 578-583.

14. Estes, M. K., Graham, D. Y., Smith, E. M. & Gerba, C.P. (1979). Rotavirus stability and inactivation. J GenVirol 43, 403-409.

15. Feinstone, S. M., Moritsugu, Y., Shih, J. W., Gerin, J. L.& Purcell, R. H. (1978). Characterisation of hepatitis Avirus. In Viral Hepatitis, pp. 41-48. Edited by G. N.Vyas, S. N. Cohen & R. Schmid. Philadelphia, PA:Franklin Institute Press.

16. Fields, B. N. & Knipe, D. M., eds. (1991). FundamentalVirology, 2nd edn. New York: Raven Press.

17. Foy, H. M. (1997). Adenoviruses. In Viral Infections ofHumans Epidemiology and Control, 4th edn., pp. 119-138. Edited by A. S. Evans & R. A. Kaslow. New York:Plenum Medical Book.

18. Frosner, G. G., Papaevagelou, G., Butler, R., Iwarson, S.,Lindholm, A., Courouce-Pauty A., Haas, H. andDeinhardt, F. (1979). Antibody against hepatitis A virusin seven European countries: a comparison of prevalencedata in different age groups. Am JEpidemiol 110, 63-69.

19. Hadler, S. (1992, June). Hepatitis A: changing epidemi-ology and the need for a vaccine. Virus and Life 14-16.

20. Halliday, M. L., Kang, L. Y., Zhou, T. K., Hu, M. D., Pan,Q. C., Fu, T. Y., Huang, Y. S. & Hu, S. L. (1991). Anepidemic of hepatitis A attributable to the ingestion ofraw clams in Shanghai, China. J Infect Dis 164, 852-859.

21. Havens, W. P. Jr. (1945). Properties of the etiologicagent of infectious hepatitis. Proc Soc Exp Biol Med 58,203-204.

22. Hermann, J. E. & Cliver, D. O. (1973). Enterovirus per-sistence in sausage and ground beef. J Milk FoodTechno! 36, 426-428.

23. Hermann, J. E., Hudson, R. W., Blacklow, H. R. &Cubitt, W. D. (1987). Marin county agent: an astrovirus.Lancet ii, 743.

24. International Committee for Microbial Safety of Food.(1996). Micro-Organisms in Foods. Vol. 5, Microbio-logical Specifications of Food Pathogens. London:Chapman & Hall.

25. Kantor, M. A. & Potter, N. N. (1975). Persistence ofechovirus and poliovirus in fermented sausage. J Food5^40,491^95.

26. Keswick, B. H., Satterwhite, N. R., Cukor, G. C.,DuPont, H. L., Secor, S. L., Bitsure, J. A., Gary, G. W.& Hoff, J. C. (1985). Inactivation of Norwalk virus indrinking water by chlorine. Appl Environ Microbiol 50,261-264.

27. LeBaron, C. W., Furutan, N. P., Lew, J. F., Allen, J. R.,Gouvea, V., Moe, C. & Monroe, S. S. (1990). Viralagents of gastroenteritis public health importance andoutbreak management. MMWR 39, 1-24.

28. Margolis, H. S., Alter, M. J. & Hadler, S. (1997). Viralhepatitis. In Viral Infections of Humans: Epidemiologyand Control, 4th edn., pp. 363-41 8. Edited by A. S. Evans& R. A. Kaslow. New York: Plenum Medical Book.

29. McKercher, P. D., Hess, W. R. & Hamdy, F. (1978).Residual viruses in pork products. Appl EnvironMicrobiol 35, 142-145.

30. Millard, J., Appleton, H. & Parry, J. V. (1987). Studieson the ehat inactivation of hepatitis A virus with specialreference to shellfish. Epidemiol Infect 98, 397-414.

31. Niu, M. T., Polish, L. B., Robertson, B. H., Khanna, B.K., Woodruff, B. A., Shapiro, C. N., Miller, M. A.,Smith, J. D., Gedrose, J. K. and Alter, M. J. (1992).Multistate outbreak of hepatitis A associated with frozenstrawberries. J Infect Dis 166, 518-524.

32. Nout, M. J. R., Rombouts, F. M. & Havelaar, A. (1989).Effect of accelerated natural lactic fermentation of in-fant food ingredients on some pathogenic organisms. IntJ Food Microbiol 8, 351-361.

33. Oishi, L, Yamazaki, K., Kimoto, T., Minekawa, A.,Utagawa, E., Yamazaki, T., Inouye, S., Grohman, G. S.,Monroe, S. S., Stine, S. E., Carcamo, C., Ando, T. &Glass, R. I. (1994). A large outbreak of acute gastroen-teritis associated with astrovirus among students andteachers in Osaka Japan. J Infect Dis 170, 439-443.

34. Panina, G. F., Civardi, A., Massirio, I., Scatozza, F.,Baldini, P. & Palmia F. (1989). Survival of foot andmouth disease virus in sausage meat products (Italiansalami). Int J Food Microbiol 8, 141-148.

35. Parrino, T. A., Schreiber, D. S., Trier, J. S., Kapikian, A.Z. & Blacklow, N. R. (1977). Clinical immunity in acutegastroenteritis caused by the Norwalk agent. TV Engl JMed 297, 86-89.

36. Parry, J. V. & Mortimer, P. P. (1984). The heat sensitiv-ity of hepatitis A virus determined by a simple tissueculture method J Med Virol 14, 277-283.

37. Parwani, A. V., Saif, L. J. & Kang, S. Y. (1990). Bio-chemical characterisation of porcine enteric calicivirus:analysis of structural and non-structural viral proteins.Arch Virol 112, 41-53.

38. Peterson, D. A., Hurley, T. R. & Wolfe, L. G. (1983).Effect of chlorine treatment on infectivity of hepatitis Avirus. Appl Environ Microbiol 45, 223-227.

39. Phithakpol, B., Varanyanond, W., Reungmaneepaiton,S. & Wood, H. (1995). The Traditional FermentedFoods of Thailand. Bangkok: Institute of Food Researchand Product Development.

40. Russel, W. C., Valentine, R. A. & Pereira, H. G. (1967).The effect of heat on the anatomy of adenovirus. J GenVirol 1,509-522.

41. Scholz, E., Heinricy, U. & Flehmig, B. (1989). Acid sta-bility of hepatitis A virus. J Gen Virol 70, 2481-2485.

42. Shirley, J. A. & Beards, G. M. (1981). The influence ofdivalent cations on the stability of human rotavirus.Arch Virol 67, 1-9.

43. Siegl, G., Weitz, M. & Kronauer, G. (1984). Stability ofhepatitis A virus. Intervirology 22, 218-226.

44. Slomka, M. J. & Appleton, H. (1998). Feline calicivirusas a model system for heat inactivation studies of smallround structured viruses in shellfish. Epidemiol Infect121,401^07.

45. Sobsey, M. (1998). Surface inactivation of hepatitis Avirus on strawberries using chlorine. Paper presented atthe 85th annual meeting of the International Associationof Milk, Food and Environmental Sanitarians, October1998, Nashville TN.

46. Sobsey, M. D., Shields, P. A., Hauchman, F. S., Davis,A. L., Rullman, V. A. & Bosch, A. (1988). Survival andpersistence of hepatitis A virus in environmentalsamples. In Viral Hepatitis and Liver Disease, pp. 121—124. Edited by A. J. Zuckerman. New York: Alan RLiss.

47. Tan, Y. W., Wang, J. X., Xu, Z. Y., Guo, Y. F., Qian, W.H. and Xu, J. X. (1991). A serologically confirmed,case-control study, of a large outbreak of hepatitis A inChina, associated with consumption of clams. Epidemi-ology & Infection 107, 651-657.

48. Wheeler, J. G., Sethi, D., Cowden, J., Wall, P. G.,Rodrigues, L. C., Tompkins, D. S., Hudson, M. J. &Roderick, P. J. (1999). Study of infectious intestinal dis-ease in England: rates in the community, presenting togeneral practice, and reported to national surveillance.Br Med J 318, 1046-1050.

49. Willcocks, M. M. , Laidler, F., Carter, M. J. & Madeley,C. R. (1990). Cultivation of human astrovirus in a con-tinuous cell line. Arch Virol 1 13, 73-82.

50. Wood, G. W. & Adams, M. R. (1992). Effect of acidifi-cation, bacterial fermentation and temperature on thesurvival of rotavirus in a model weaning food. J FoodProt 55, 52-55.

CHAPTER 9

Microbiological Hazards andTheir Control: Parasites

Mike Taylor

INTRODUCTION

Many types of parasites are food-borne, andhumans can become infected following the in-gestion of infected or contaminated meat, fish,molluscs, vegetables, or fruit, or products de-rived from these foods. In most cases, parasiticinfections are acquired by eating raw or incom-pletely cooked food, or food that is partiallypickled or smoked or poorly preserved. Most, ifnot all, infections are preventable if the food isprepared sufficiently to destroy the infectivestages of the parasite. However, many infectionsare commonly associated with cultural and eat-ing habits that have been in practice in popula-tions for generations.

Meat from many species of animals has beena recognized source of many helminth, and someprotozoal, infections, in man. In developedcountries, the introduction of meat hygiene mea-sures has resulted in reduced incidences of manyof the traditionally recognized helminth infec-tions. However, the eating of many traditionalraw or lightly cooked meat dishes continues, andmay occasionally result in infection in man.Raw, uncooked fish dishes are also commonlyeaten in many cultures, and snails, clams, oys-ters, and a variety of other molluscs are part ofthe diets of many people worldwide. Althoughmost food-borne helminth infections are re-ported from Third World countries, increased

Source: © Crown copyright 1997. Published withthe permission of the Controller of Her BritannicMajesty's Stationery Office. The views expressed arethose of the author and do not necessarily reflect thoseof Her Britannic Majesty's Stationery Office or theVLA or any other government department.

immigration, tourism, and desire to experiencethe culinary dishes of other cultures may in-crease the incidence of parasitic disease in othercountries. As with some helminths, protozoanparasites are opportunistic infections that are of-ten acquired as the result of poor hygiene ortravel to foreign countries.

Fermented foods that use raw ingredients thathave been contaminated or infected with infectiveor intermediate parasite stages have the potentialto cause human infection. Although food-borneparasitic diseases continue to be reported globally,reports of human infection following the ingestionof fermented food are sparse. Cultural dishes suchas som fak (a fermented Thai minced fish dish)have been reported to result in human infectionwith the helminth parasite, Gnathostoma (seeGnathostomosis). Infection with another proto-zoan parasite, Giardia, has been reported follow-ing the ingestion of cheese dip. Fermentationalone may therefore be insufficient to prevent thetransmission of many food-borne parasites, andpotentially infected material should be avoidedwherever possible, or alternatively subjected tofreezing or some form of heat treatment.

Many parasites infect humans, but referenceis made only to those helminths and protozoathat are recognized as food-borne and a potentialsource of infection in fermented foods.

NEMATODES

The nematodes are a diverse group of para-sitic or free-living unsegmented worms that areusually cylindrical and elongate in shape. Withfew exceptions, the sexes of nematodes are sepa-rate, and the life cycle may be direct or indirect,

involving an intermediate host. The general de-scription and classification of the food-bornenematodes discussed in this chapter are summa-rized in Table 9-1.

Angiostrongylosis

Angiostrongylus cantonensis and A. costari-censis are parasites of rodents (predominantlyrats) that can cause disease in humans through theingestion of infected snails. Two other species, A.malaysiensis and A. mackerrasse, occur but havenot been associated with human infection. Mosthuman infections are acquired by eating infectedsnails, but can occur through eating infectedshrimp and crab, or snail- or slug-contaminatedraw vegetables.

Public Health Significance

A. cantonensis can cause meningitis and menin-goencephalitis with mild to moderate symptoms,often of sudden onset, with intense headaches;vomiting; moderate intermittent fever; and, in ap-proximately 50% of cases, coughing, anorexia,malaise, constipation, and somnolence.176 In se-vere cases, coma and death may occur. Accidentalingestion of the slug intermediate host of A.costaricensis causes abdominal angiostrongylosiswith symptoms similar to appendicitis, includingfever, abdominal pain, anorexia, diarrhea, andvomiting.96 Migratory larvae may cause gastroen-teritis, tumor-like masses or abcessation of the in-testines, liver enlargement, nervous signs, coma,and, occasionally, death.108 Infection is normallydiagnosed by confirming the presence of parasitesor eggs in surgically removed tissues or fluids orby using serological assays.37'144'169

Life Cycle

Adult worms are found in the pulmonary ar-teries of rats. The life cycles of A. cantonensisand A. costaricensis are similar and are shown inFigure 9-1 . The prepatent period (from infectionto maturity) is 42—45 days.

Distribution

The distributions of A. cantonensis and A.costaricensis are summarized in Table 9-2.

Most human infections and deaths associatedwith A. cantonensis have been reported fromTaiwan, Thailand,131'176 and some Pacific is-lands, but human infections have been reportedin most countries where the parasite occurs andappears to be spreading.33 The incidence ofangiostrongylosis appears to be spreading, espe-cially in those areas in which snails are an im-portant part of the diet.

Epidemiology and Transmission

Humans become infected with angiostrongyl-osis by intentionally or accidentally eating in-fected snails or slugs. Rodents are infected by in-gesting infected molluscs or by ingesting infectivelarvae present in "slime" on plants. In most en-demic areas of Asia, both land and aquatic snails(Achatina and PiId) are eaten regularly. The giantAfrican land snail, Achatina fulica, is a particulardelicacy in many countries and is a good interme-diate host. Slugs and snails (which are used formedicinal purposes in some cultures), land crabs,shrimp, and paratenic hosts such as toads andfrogs have also transmitted infection. Sauces pre-pared from shrimp juices or unwashed contami-nated vegetables have also been incriminated. Thedrinking of untreated water containing larvae re-leased from dead snails has also been suggested asa means of infection. Infection continues to be re-ported from new areas of the world, in part due tothe dissemination of the intermediate snail hosts,but also due to transportation of infected rats onships. Most terrestrial and aquatic snails are sus-ceptible to infection, and populations can bereadily infected from carrier rats.

Prevention and Control

Angiostrongyliosis can be prevented by edu-cating people in endemic areas to avoid eatinguncooked molluscs, particularly land snails.Freezing will kill larvae present in snails if theyare frozen at -15 0C for 12-24 hours. Paratenichosts (i.e., shrimp, prawn, crabs) should becooked before eating, and vegetables should bewashed before eating raw. Little or no informa-tion is available on the survivability of infectivelarvae in fermented foods such as balao-balao(fermented shrimp).

Table 9-1 Food-Borne Helminths— Classification

Genus

Angiostrongylus

AnisakisTrichinellaCapillariaDioctophymaGongylonemaGnathostomaDiphyllobothrium

Taenia

FasciolaFasciolopsisEchinostomaHypoderaeumParagonimusNanophyetesHeterophyesMetagonimusOpisthorcis(Clonorchis)

Family

Metastrongyloidae

AnisakidaeTrichinellidaeCapillaridaeDioctophymatidaeThelaziidaeGnathostomatidaeDiphyllobothridae

Taeniidae

Fasciolidae

Echinostomatidae

Troglotrematidae

Heterophyidae

Opisthorciidae

Order

Strongylida

AscarididaEnophida

DioctophymidaSpirurida

Pseudophyllidea

Cyclophyllidea

Echinostomida

Plagiorchiida

Opisthorchiidae

Class

Nematoda (Roundworms)Elongate, cylindrical, unsegmented worms withfluid-filled body cavity. Sexes are separate.Life cycle direct or indirect.

Cestoda (Tapeworms)Tape-like segmented body comprising head,neck, and strobila (proglottids). Bothria or suckersfor attachment. Hermaphroditic. Indirect life cycles.

Trematoda (Flukes)Unsegmented leaf or lancet-shaped worms withtwo muscular suckers for attachment and well-developed oral sucker and pharynx.Hermaphroditic (generally). Indirect life cycles.

Phylum

Nemathelminthes

Platyhelminthes

Source: ©Crown Copyright.

Anisakiosis

Anisakiosis is a parasitic gastrointestinal dis-ease of man that is caused by the larval stages ofnematodes of the family Anisakidae. The adult

worms are common parasites of marine mam-mals (Table 9-2), and the larval stages are foundin marine fish and squid. Humans acquire infec-tions by eating the raw or improperly cooked orpreserved meat of these animals. There are many

Figure 9-1 Life cycle of Angiostrongylus cantonensis. Source: © Crown Copyright.

Human infection through ingestion ofraw or undercooked intermediate host(snails) or carriers (crabs, prawns, etc.)

Carriers(crabs, prawns)

Second- and third-stage larvae inintermediate host (slug, snail)

Angiostrongylus cantonensis

Ingested by rat

First-stage larvaeexcreted in feces

Table 9-2 Food-Borne Helminths (Nematodes, Cestodes)

Source of Infection to Man

Molluscs (snails), shrimps,crabs, amphibians,contaminated vegetablesand salads

Fish

Fish

Fish

Fish , frogs, chickens,ducks, snakes

Salads (insects)

Meat

Fish

Meat (beef)

Meat (pork)

Main Hosts (final)

Rodents (rats)

Pinipeds (whales, dolphins,porpoises)

Seals, sealions, walrusesMan (several species of

birds)Carnivores (mink, ferret,

dogs, cats, jackals), manCarnivores (dogs and

cats), manPigs, manRuminants, pigs, dogs,

cats, horses, rodents,primates, man

Pigs, rodents, carnivores(mink, fox, badger,bears, walrus, seals),man

Dog, fox, mink, cat, pig,bear, seals

Man

Man

Distribution

Asia and Pacific Islands, Australia, India,Africa, Caribbean, parts of UnitedStates

North and South America, Pacific Islands,parts of northern Europe

Phillipines, Thailand

Worldwide (North and South America,southern Europe, Asia, Middle East)

Thailand, Japan, southeast Asia, China,Mexico (Middle East, Africa, BalticStates, Russia)

United States, former USSR, parts ofEurope, Middle East, China, NorthAfrica, New Zealand

Worldwide (except Antartica)

Northern Hemisphere (northern Europe,Russia, North America, South America,Asia, Africa)

Central Africa, Asia, South and WestAfrica, parts of Europe, southeast Asia,Central and South America; alsoreported in United States, Canada,Australia and Pacific Islands

Central and South America, central andeast Africa, southeast Asia, southernEurope

Parasite

Angiostrongylus

Anisakis simplexAnisakis decipens

Capillaria phillipinensis

Dioctophyma renale

Gnathostoma spinigerum

Gnathostoma hispidumGongylonema pulchrum

Trichinella spiralis

Diphyllobothrium latum

Taenia saginata(Cysticercus bovis)

Taenia solium(Cysticercuscellulosae)

Parasite Class

Nematoda(Roundworms)

Cestoda(Tapeworms)

Source: © Crown Copyright.

species of anisakid nematodes, but those mostoften associated with human illness are Anisakissimplex, Pseudoterranova decipiens, Phoca-nema, and Contracaecum spp.


Anisakis causes acute or chronic gastrointesti-nal disease in man. Migrating larvae cause a for-eign-body reaction and eventually necrosis andhemorrhage of the stomach, occasionally creat-ing tunnels and burrows in the stomach mu-cosa,139 causing pain, nausea, and vomiting. Theacute symptoms subside in a few days, but avague abdominal pain with intermittent nauseaand vomiting persists for weeks, with symptomsresembling those of a peptic ulcer. The conditionis often misdiagnosed because of its similarity toother acute gastrointestinal conditions (e.g., gas-tric ulcers or neoplasm, appendicitis, diverticuli-tis, Crohn's disease, gallstones, etc.). Diagnosisis normally only made following biopsy andconfirmed on histopathology.

Life Cycle

Adult anisakid worms are located in thestomach of marine cetaceans and pinnipeds butdo not develop in humans. The life cycle ofthese worms is shown in Figure 9-2. Humansbecome infected if an intermediate fish hostcontaining infective third-stage larvae is eatenuncooked.

Distribution

Fish infections are found in most oceans andseas but are highest in areas in which there arehigh marine mammal populations, such ascoastal Japan and Alaska. Many species of fishare naturally infected with anisakiosis, and theprevalence of infection can be very high. InJapanese waters, 123 species of marine fish havebeen found to harbor the parasite.118 The globaldistribution of anisakiosis is summarized inTable 9-2. The highest levels of infection havebeen reported from Japan and the Netherlands.88


The main source of infection from anisakidworms for man is marine fish, many species ofwhich are highly parasitized by anisakid larvae.

Humans acquire A. simplex by eating raw orpoorly salted, pickled, or smoked herring, cod,mackerel, salmon, or squid, and P. decipiensfrom cod, halibut, flatfish, and red snapper.Such traditional preparations as green herring,lomi lomi salmon, seviche, sushi, and sashimi(i.e., seasoned fish fillets), all of which use rawor uncooked fish, are major sources of infec-tions.34 In the Netherlands, the occurrence of thedisease was due to the habit of consuming rawor lightly salted herring (green herring).162 Al-though the habit persists, the incidence of hu-man anisakiosis has been drastically reduced byfreezing fish before marketing. In recent years,the highest incidence of the disease has been re-corded in Japan, where many fish dishes areeaten raw (sashimi); pickled in vinegar(sunomono)\ or fermented in rice, rice bran, orkoji (sushi and zuke). In the United States, atleast two cases have been linked to eatingseviche (i.e., pieces of raw fish seasoned inlemon juice for several hours), and others to eat-ing Japanese raw fish dishes.


The risk of human infection with anisakiosisincreases in countries where fish are eatenraw.172 Anisakiosis is preventable by ensuringthat only well-cooked marine fish, octopus, andsquid are eaten. Larvae are killed by cooking at60 0C or above. Freezing fish at -20 0C for 24hours will kill the larvae, with the exception ofsome North American species that can survivefreezing at that temperature for 52 hours. Clean-ing and eviscerating fish immediately after theyare caught prevents larvae migrating from theintestine to the muscles. Salt curing, marinating,microwaving, and smoking temperatures are in-sufficient to kill the parasite.34 Marination ofherring has been a long tradition in parts ofnorthern Europe,85'106 involving the preservationof herring fillets in salt and acetic acid. The salt/acetic acid marination process produces the typi-cal flavor as a result of denaturing of the fishproteins, lowering of the pH, and addition ofsugar and spices.158 The salt/acid treatment maytake up to 42 days to kill Anisakis larvae, and atlow concentrations may not kill larvae at all.85

Larvae have also been shown to survive in izushi

Figure 9-2 Life cycle of Anisakis simplex. Source: © Crown Copyright.

(i.e., pickled rice with cod roe, fillets of salmon,and cod) and in sashimi.118

Capillariosis

Intestinal capillariosis is caused by a tinynematode, Capillaria philippinensis. The dis-ease was first recognized in the Philippines inthe 1960s and has subsequently been reported in

other countries.32'34'93 Man is the main definitivehost for the parasite, but several species of birdsare now believed to be a natural host and are ableto transmit the infection.35


Infections with C. philippinensis can cause di-arrhea, anorexia, weight loss, and, if left un-treated, death. The parasites cause damage to the

Anisakis simplex

Eggs in feces

First- or second-stage larvaeingested by crustacean

Infective larvae in fish

Human infection through ingestionof raw, undercooked, or poorlypreserved fish

mucosa of the small intestine, leading to fluid,protein, and electrolyte loss. Clinical symptomsinclude abdominal pain and diarrhea. If treat-ment is not initiated rapidly, patients die becauseof the irreversible effects of the electrolyte loss,heart failure, or septicemia. In endemic areas, a di-agnosis can be made on patients based on clinicalsigns. In chronic infections, there is weight loss,wasting, and intractable diarrhea. Infection is con-firmed by identifying eggs, larvae, and adults mi-croscopically in the feces. Serological tests are notsufficiently specific for routine use.

Life Cycle

C. philippinensis is a small nematode that isfound in the small intestines of man. Further de-velopment occurs in fish. A number of species offreshwater fish are suspected to serve as inter-mediate hosts for the parasite. The fish are usu-ally small and are eaten whole in the Philippinesand Thailand, leading to human infection.

Distribution

The distribution of intestinal capillariosis issummarized in Table 9-2. Nearly 2,000 cases ofcapillariosis have been recorded in the Philip-pines, with more than 100 deaths. In Thailand,there have been only a few hundred cases, withan unknown number of deaths recorded.36


Sanitation facilities are poor in many rural ar-eas of southeast Asia and defecation in the fieldsis a common practice. During the monsoonrains, the feces are broken down and washedaway to streams and ponds, resulting in infectionof freshwater fish. These and many other foodsare eaten uncooked, especially in the endemicareas of the Philippines and in Thailand.


Educating both the local population and visi-tors of the dangers of eating uncooked freshwa-ter fish can help prevent intestinal capillariosis.Improvement of sanitation and the control of in-discriminate disposal of feces would also bebeneficial. No information is available on theconditions that kill the parasite in fish and

whether they are known to survive in fermentedfish products.

Dioctophymosis

Dioctophyma (Dioctophyme) renale is a largekidney-dwelling nematode of carnivores. Infec-tion is acquired by ingesting raw or undercookedfish or frogs. It has only rarely been reported inman.


Infection in man can cause renal damage andassociated symptoms of renal colic, hematuria,or urinary obstruction. In humans and dogs, Dio-ctophyma usually locates in only one kidney,most often the right one, and in most cases onlyone parasite is found, causing little or no clinicalsymptoms. Infection is diagnosed by the pres-ence of the characteristic thick, bipolar eggs inurine (only if female worms are present).

Life Cycle

D. renale is a large, blood-red nematode that isfound in the kidneys of carnivores (Table 9-2).The size of the parasite depends on the size of thehost species; in dogs, the adult female of the para-site can reach up to 1 m in length. The intermediatehost is a free-living, aquatic segmented worm (oli-gochaete) in which further development to an in-fective larva occurs. Transmission to a final hostoccurs following ingestion of the annelid interme-diate host, or more usually following ingestion of afish or a frog, which act as transport hosts. Theprepatent period is from 3]/2 to 6 months.

Distribution

D. renale is found on all continents with thepossible exception of Africa and Oceania (Table9-2). The most frequently reported form of in-fection is canine dioctophymosis, with the high-est prevalence of infection in Canadian wildmink (Mustela vison), where 18% of the animalsare infected. The disease is very rare in man,with only a few reports worldwide.


In North America, mustelids, especially mink,appear to be the main reservoir. In other areas, it

is likely that other species of mustelids or wildcanids serve as main definitive hosts. Thesehosts are infected by ingesting frogs or fish(paratenic hosts) and aquatic oligochaetes (inter-mediate hosts) that contain the third-stage lar-vae. Dogs and humans are accidental hosts andare infected by ingesting raw fish and frogs, andalmost always harbor only one parasite. The rar-ity of human infection can be explained by thefact that larvae are located in the mesentery andliver of fish and frogs, which generally are noteaten by man.


Infection can be prevented, both in humansand dogs, by avoiding the consumption of raw orundercooked fish and frogs. It is not known whatconditions result in larval death.

Gnathostomosis

Gnathostomosis is caused by infection withthe larval or immature adult stages of nematodesof the genus Gnathostoma. Adult parasites arereported in dogs, cats, and other carnivorous ani-mals worldwide. In Thailand, more than 40 spe-cies of vertebrates have been reported to be natu-rally infected. These include freshwater fish,frogs, snakes, chickens, ducks and other birds,rats, mongooses, and tree shrews.38 Four species,G. spinigerum and G. hispidum, and more re-cently, G. doloresi and G. nipponicum, havebeen reported in humans in Japan.111'152


Man is an abnormal host, with infection re-sulting in a larval migrans causing red, itchy,and edematous subcutaneous swellings that usu-ally last approximately one week but can recurweeks or months later. More rarely, the parasitemay enter the eye, causing subconjunctivaledema, exopthalmus, impaired vision or blind-ness through hemorrhage, and retinal damage.Invasion of the central nervous system (CNS)can produce headaches, neck stiffness, drowsi-ness, or coma and death. Brain hemorrhage andtransitory obstructive hydrocephalus have alsobeen reported.140 Diagnosis in endemic areas is

based on history, symptoms, or serology and canbe confirmed following the recovery and identi-fication of parasitic larvae.114

Life Cycle

Adult worms of G. spinigerum are found intumor-like masses in the stomach wall of fish-eating mammals. The life cycle of G. spinigerumis shown in Figure 9-3. The prepatent period isapproximately six months.

Distribution

The distribution of human gnathostomiosis issummarized in Table 9-2.


The main source of human infection ofgnathostomosis in Thailand is the snake-headedfish, Ophicephalus spp., which is one of thefishes used in som-fak, a rice-fermented fish dishwith widespread popularity. In Japan, freshwa-ter Ophicephalus species are eaten raw assashimi. The ingestion of raw or inadequatelycooked fish is the major source of infection inother areas reporting the disease. Infections inMexico are attributed to eating raw cycloid fish asceviche." Human infections are also reportedfrom eating raw or poorly cooked catfish, eels,frogs, chickens, ducks, and snakes.39 Dogs, cats,and several species of wild mammals are reser-voirs of the parasite. These definitive hosts be-come infected primarily through eating infectedfish or other animals that serve as paratenic hosts.


Health education programs in endemic areasof Asia are required to control this type of infec-tion. Ensuring that people eat only well-cookedfish, eels, or other intermediate hosts such assnakes, frogs, and poultry can prevent infec-tions. Potentially copepod-infested water shouldbe boiled or treated.

Gongylonemosis

Gongylonema pulchrum is a spiruroid nema-tode of the Thelaziidae family (Table 9-1). It isfound in all domesticated mammals, but is most

Figure 9-3 Life cycle of Gnathostoma spinigerum, Source: © Crown Copyright.

prevalent in ruminants (Table 9-2). Cases of hu- palate, soft palate, and tonsils, with pharyngitisman infection are rare. and stomatitis reported. Diagnosis is based on

history, clinical signs of mouth irritation, andPublic Health Significance microscopic identification of the parasite. Occa-

In humans, gongylonemosis parasites have sionally, the parasite has been found emergingbeen found in the submucosa of lips, gums, hard from the mouth. Infection in animals is usually

Human infection throughingestion of raw orundercooked fish

Encysted larvae in fish,amphibians, or reptiles

Cyclops, crustaceanintermediate host

Gnathostoma spinigerum

Hatched larvaeingested by Cyclops

Final Host(fish-eating mammals)

Eggs passedin feces

Adults worms in stomach

asymptomatic, but may sometimes cause lesionsof the mouth or pharynx.29 In pigs, the parasite isfound in the tongue mucosa and may cause oc-clusion of the esophagus.179

Life Cycle

Adult Gongylonemosis worms live in themouth, esophagus, or rumen of the final hosts. Thelife cycle of G. pulchrum involves coprophagicbeetles of the genera Aphodius, Ontophagus, andBlaps as intermediate hosts. Ruminants becomeinfected by ingestion of the beetles with grass orother infested food.

Distribution

Although G. pulchrum is widely distributedgeographically, human infection is rare. The dis-tribution of G. pulchrum is summarized in Table9-2. Infection in domestic ruminants varies con-siderably, with high levels of infection reportedin the Ukraine (e.g., 32-94% of adult cattle, 39-95% of sheep, and 0-37% of pigs infected),29

and Iran (e.g., 49.7% of the cattle).4 In theUnited States, the parasite was found in 5.9% ofpigs, varying from O to 21% according to geo-graphic origin.179


Man is an accidental host and is probably in-fected by ingesting beetles on salads and rawvegetables.


Because the Gongylonemosis parasite is rarein humans and causes only mild symptoms, spe-cial control measures are usually not recom-mended. Individual protection can be obtainedby observing the rules of personal, food, and en-vironmental hygiene.

Trichinellosis (Trichinosis)

Trichinellosis (trichinosis) is a food-bornedisease that is caused by infection with parasiticnematodes of the genus, Trichinella. Infectionresults from ingesting meat harboring infectivelarvae. Until recently, T. spiralis was acceptedas the sole representative of the genus Tri-chinella. It has now become clear that not all

populations of the parasite are the same, and fivespecies have recently been proposed.24 T.spiralis is the most important species, withwidespread distribution in domestic pigs.


Trichinellosis is a cosmopolitan zoonotic,food-borne, parasitic disease resulting from theingestion of meat harboring the infective larvaeof Trichinella. Intestinal trichinellosis is usuallymild and therefore not often diagnosed, but maycause diarrhea and abdominal discomfort, some-times accompanied by nausea and vomiting,leavy infections can be severe, causing ex-

tremely painful muscles, and even fatalities inman. The illness typically begins one to fourweeks after ingestion of infective meat, and ischaracterized by muscle aches and fever. Musclepain, which may be severe and incapacitating, isan outstanding feature, with the jaw muscles fre-quently involved. The infection of heart musclemay induce severe, even fatal, myocarditis.Death may also result from encephalitis or pneu-monitis. In most parts of the world, the probabil-ity of contracting the disease is now low, and itsdecline, certainly throughout the western world,can be attributed to the introduction of a numberof control measures discussed later. Diagnosismay be based on a clinical history of prior con-sumption of raw or undercooked meat, but thisaspect may be missed, particularly in countrieswhere the disease rarely occurs. Clinical signs ofmuscle aches, fever, and periorbital edema areindicative of the disease. Infection is usually di-agnosed by demonstrating larvae in muscle bi-opsy specimens.

Life Cycle

T. spiralis is a small, nematode parasite that isfound in the small intestine of man, pig, rat, andother mammals (Table 9-2). The life cycle of T.spiralis is shown in Figure 9-4. Encysted larvaeremain infective for months, even years. In somehosts, especially man, they eventually die andbecome calcified. Trichinella is remarkableamong parasitic nematodes in having neither afree-living stage between individual hosts nor anintermediate host. The spread of larvae through-out the host musculature permits transmission to

Figure 9-4 Life cycle of Trichinella spiralis. Source: © Crown Copyright.

another individual host, but only after the deathof the first host.

Distribution

Trichinella has been recorded on every conti-nent except Antarctica (Table 9-2). In many ar-

eas, its prevalence in man and pigs has been re-duced to a low level, but persists at higher levelsin wildlife. Prevalence rates are higher in largecarnivorous or scavenging species such as bears,mink, fox, badger, wild boar, and rodents. Infor-mation on the global incidence of clinical

Sylvatic cycleGame meat

Wildlife infection fromcarrion feeding

Trichinella spiralis

Infection via deadinfected rats

Rats infected frompork scraps

Pig-to-pig transmission viacannibalism, tail-biting, etc.

Domestic cycle

Jluman infection via raw oriercooked pork, sausages, etc.

Human infectionfrom game meat

trichinellosis in man is not readily available, butit is evident that there has been a striking reduc-tion in western Europe and the United States.Several outbreaks in France and Italy have beenassociated with eating horse meat that wasthought to have become infected by intention-ally or inadvertently eating fodder that was con-taminated with the bodies of dead rats or mice.Outbreaks of varying severity continue to occurin eastern Europe, the former Soviet Union,Asia, and occasionally elsewhere.


Two distinctive epidemiological cycles of T.spiralis are recognized, domestic and sylvatic.The domestic cycle involves pigs and rats,through either the feeding of uncooked porkscraps (e.g., in waste feed) to pigs, cannibalismamong pigs (e.g., scavenging of dead carcassesand possibly tail biting of live pigs), or infectionof rats by disposal of uncooked pork and the sub-sequent infection of pigs by the ingestion of in-fected rats. Human infection occurs whenundercooked or raw pig meat is eaten.

In the sylvatic cycle, Trichinella is transmit-ted among species of wildlife. Man becomes in-fected by consuming game meat. The two cyclesmay be interrelated by the fact that domesticpigs may scavenge on dead wild animals andvice versa. Rats can also be of significance inthis respect in that they may become infected ei-ther through eating pork scraps or by scavengingwildlife carcasses. Rats may in turn be huntedand eaten by wildlife predators, or when theydie, be eaten by wild carrion feeders or pigs.Crossover of Trichinella from one transmissioncycle to another may occur, with domestic tosylvatic cycle the more likely109 as wildlifestrains may have low infectivity for rats or pigs.


Human infection by trichinae is linked to theconsumption of undercooked pork or gamemeat. A number of measures can therefore betaken to control infection. These involve a com-bination of consumer education and control ofinfection in pigs. Pig herds can be kept free ofinfection by ensuring that pork scraps are not fed

to pigs. In some countries, laws requiring wastefood (swill) to be cooked before being fed topigs have been introduced to help control a num-ber of important viral pig diseases. Good hus-bandry and rat-proofing piggeries also help con-trol infection. In many countries, meatinspection procedures have been introduced toidentify infection.

Ethnic or cultural practices in which meat iseaten undercooked or raw are a particular riskarea and may call for special educational mea-sures. Traditional types of sausage (includingfermentated sausage) that receive little cookinghave long been associated with human disease.Education draws attention to the need to freezeor cook such foodstuffs. Freezing may be moreacceptable because it has less effect on flavorand taste than cooking. To be effective, meatmust be frozen throughout and stored for suffi-cient time to ensure death of encysted larvae.The thermal death point of T. spiralis is approxi-mately 57 0C, and cooking pork so that it reachesa temperature of 77 0C gives a margin of safetywithout destroying the taste of the meat. Slightlylower temperatures may be adequate providedthe temperature is achieved uniformly through-out the meat. It can be assumed that pork will besafe if it is cooked until there is no red or pink col-oration throughout. In the United States, fer-mented pork sausages must either originate fromTrichinella-CQrtified meat or be heated to 58.3 0Cat the end of the fermentation process.89 Althoughcertain factors such as pH may affect survivabilityof trichinae in fermented sausage, prescribed heattreatment is considered necessary.31

CESTODES (TAPEWORMS)

There are a number of food-borne tapewormsthat infect humans, of which a few are acquiredby eating meat or fish. A general description andclassification of food-borne tapeworm species isprovided in Table 9-1.

Diphyllobothriosis

Although there are various species of fishtapeworm reported in humans, the most impor-tant is Diphyllobothrium latum.


Diphyllobothriosis is usually asymptomatic,but long-term infection with the tapeworm mayproduce vitamin B 12 deficiency leading to anemia.The presence of several worms may cause intesti-nal obstruction. CNS signs of peripheral and spi-nal nerve degeneration have been reported. Diag-nosis is based on identifying the presence of thecharacteristic eggs in human feces.

Life Cycle

The fish tapeworms are among the largestworms to infect humans. Adult tapeworms arefound in the small intestine, ranging from 2 m to15 m in length and living as long as 10 years.Many species offish such as pike, perch, turbot,salmon, and trout serve as second intermediatehosts. The life cycle of D. latum is shown in Fig-ure 9-5.

Distribution

The distribution of D. latum is widespread inthe temperate and subarctic regions of the North-ern Hemisphere where freshwater fish are eaten.Distribution is summarized in Table 9-2.


Diphyllobothriosis is transmitted by the in-gestion of infected, raw, or improperly cookedfreshwater fish. Human cultures with prefer-ences for smoked, pickled, or raw fish (such assashimi or sushi) are at particular risk.

Endemnicity is maintained where there ispoor sanitation, or, due to the practice of lake-side hotels, pumping of raw sewage into fresh-water lakes. The parasite may have been spreadby the emigration of infected people from en-demic areas into lake regions in which suscep-tible copepods and fish were present. Scandina-vian and Russian immigrants are thought to haveintroduced the parasite into the Great Lakes re-gion of North America and Alaska.


Human infection is easily prevented by ensur-ing that freshwater fish is well cooked or bychanging dietary habits through education. Iffish is to be eaten raw, smoked, pickled, or fer-

mented, then freezing at -10 0C for 1-3 daysshould kill pleurocercoid larvae in the fish tis-sue. Sewage from lakeside hotels and from lei-sure boats should be treated before release intothe lakes.

Taeniosis (Cysticercosis)

Taeniosis is intestinal infection with a pre-adult or adult stage of a tapeworm of the genusTaenia. Only two species, T. saginata and T.solium, infect human intestines as adult tape-worms, and man is the only known final host.Intermediate stages of these human tapewormsinfect the tissues of meat-producing animals.Tissue infection with a metacestode stage ofTaenia is referred to as cysticercosis. The inter-mediate hosts of T. saginata and T. solium aredomesticated cattle and pigs, respectively, butseveral other ruminants, including sheep, goats,and llamas, have been recorded as carriers of C.bovis, although the validity of these hosts hasbeen questioned. The reindeer, Rangifertarandus, has been shown to act as an intermedi-ate host in Russia. Cysticerci thought to be C.cellulosae have been reported in monkeys, wildpigs, bush babies, camels, rabbits, hares, bears,dogs, foxes, cats, rats, and mice. Four species(i.e., T. solium, T. saginata, T. multiceps, and T.hydatigena) parasitize human organs as cys-ticerci; the last three species invading man onlyexceptionally.163 Other species, T. serialis, T.longihamatus, T. crassiceps, and T. taeniae-formis, have occasionally been diagnosed as acause of cysticercosis in man.12'14'92


Taeniosis itself causes little if any disabilityand morbidity. Conversely, human cysticercosishas a definite public health significance, causingdisability and death in infected people. Cys-ticerci may be found in every organ of the bodyof man, but are most common in the subcutane-ous tissue, eye, and brain. In the brain(neurocysticercosis), the tissue reaction that oc-curs may cause a variety of CNS disorders, someof which may be fatal. In endemic areas,neurocysticercosis may be responsible for 30-60% of cases of epilepsy, and mortality may

Coracidium ingested by crustacean(first intermediate host)

Figure 9-5 Life cycle of Diphyllobothrium latum. Source: © Crown Copyright.

range from 1% to 2% of all causes of death. For along time, treatment of taeniosis was neither safenor satisfactorily effective; however, the intro-duction of effective taenicides has seen a dra-

matic improvement in therapy. Diagnosis of hu-man taeniosis is not always straightforward andis based on clinical symptoms and the identifica-tion of tapeworm proglottids or eggs during

Procercoid develops withinfirst intermediate host

Egg hatches in water,releasing coracidium

Gravid Proglottidin feces

First intermediate host ingestedby fish

(second intermediate host)

Diphyllobothrium latum

Adult worm insmall intestine

Human infection through ingestionof raw, uncooked, or poorlypreserved fish containingplerocercoids

stool or anal swab examination.72 Repeated fecalexaminations increase the chances of findingeggs, and diagnosis can be improved by takingperianal swabs using sticky cellophane tape. Afecal antigen test has also been reported.3'41

Life Cycle

Adult tapeworms of T. saginata measure 4-12m in length and take approximately three monthsto become fully grown. T. solium is shorter, 1.5 mto 8 m, and takes 62-72 days to mature. Bothtapeworms are found in the small intestine ofman and may live for up to 25 years. The lifecycles are similar, involving cattle and pigs re-spectively (Figure 9-6).

Distribution

T. saginata (C. bovis) occurs in many cattle-breeding regions and especially where beef iseaten raw or undercooked. Distribution of T.solium (C cellulosae) is usually confined to poorcountries because it is mainly related to the lowsanitation in pigs' breeding areas. Human migra-tion and increased consumption of pork increasesthe spread of taeniosis/cysticercosis from the en-demic rural areas into urban areas. Bovine cys-ticercosis has a high prevalence in western andeastern Central Africa and some Asian countries.60

Moderate infections are seen in other coun-tries125'173 and are summarized in Table 9-2. Theprevalence of bovine cysticercosis in Europe andthe United States has decreased throughout thetwentieth century, through the introduction ofmeat inspection procedures in these countries.Fluctuations in prevalence in Europe have oc-curred as a result of the migration of people fol-lowing World War II and increased tourism,124 andin the United States, occasional outbreaks havebeen attributed to Mexican immigrant workers.141

The distribution of pig cysticercosis is alsosummarized in Table 9-2. In countries that aresupposed to be free of T. solium infections, humancysticercosis may occur sporadically, mainly dueto contact with infected migrants or tourists.76


Man becomes infected with taeniosis by eat-ing raw or undercooked meat containing viable

cystercerci. The life span of cysticerci varies,probably due to parasite or host strain variations,different infection doses, and host immunologi-cal responses. C. bovis in cattle, for example, re-main viable for nine months to three years.Numbers of cysticerci depend not just on the de-gree of exposure to infective and viable tape-worm eggs, but also on the level of host immu-nity that can effectively reduce the number ofdeveloping cysticerci.

Adult tapeworms have a high biotic potential.Within infected human populations, the totalproduction of eggs can be enormous. The eggsare sensitive to temperatures higher than 38 0Cand to desiccation, but are capable of survivingEuropean winters for 35 days58 and of survivingin sea or brackish waters for some time. Theeggs are also relatively resistant to chemicals. T.saginata eggs remain infective for calves after16 days in sewage, up to 71 days in sludge, andfor several months in pastures.110

Animal husbandry practices, human eatinghabits, sanitary education, and willingness to co-operate in control programs are all factors thatmay further influence the transmission oftaeniosis. Poverty, ignorance, and some localcustoms of eating raw pork or raw meat or vis-cera may also play an important transmissionrole. The habit of eating raw meat or semi-rawsausages is strongly rooted in some cultures.


Control measures include improved diagnosisand treatment of human taeniosis as well asproper pig and cattle husbandry and the instiga-tion of meat inspection procedures. However,routine meat inspection may be inadequate forlight infections and may miss more than half ofthe infected animals.165 In cattle, the heart ismost frequently infected; in pigs, the tongue isthe most commonly infected organ.125 Poor localsanitation is largely responsible for the ease oftransmission in rural endemic areas.

Food transmission can be controlled by avoid-ing the consumption of raw pork or beef or semi-raw meat products. Freezing infected pork orbeef can kill the cysticerci when the internaltemperature of meat or a carcass is less than -5

Figure 9-6 Life cycle of Taenia solium and Taenia saginata. Source: © Crown Copyright.

Adult worms in smallintestine

Gravid proglottids in feces

T. saginata T. solium

proglottids

Cysticercus evaginates

Human cysticercosisfollowing ingestion ofeggs

cysticercus

Taenia soliumCysticerci in intermediate hosts

Taenia saginata

Eggs (hexacanthembryo) ingested byintermediate hosts

0C for at least four days or around -20 0C for atleast 12 hours. It is not known if cysterci can sur-vive meat fermentation conditions, but it wouldseem sensible to freeze or heat treat flesh ingre-dients wherever possible.

TREMATODES (FLUKES)

Trematodes or flukes are unsegmented,dorso-ventrally flattened, leaf- or lancet-shapedparasitic worms. The life cycles of the parasiticflukes discussed in the text are indirect, involv-ing a variety of intermediate hosts. A generaldescription of food-borne flukes and their classi-fication is provided in Table 9—1.

Clonorchiosis

Clonorchis sinensis (Opisthorchis sinensis), theChinese liver fluke, is a trematode infection that isfound throughout southeast Asia. The parasite isacquired by eating raw or poorly cooked or pre-served freshwater fish, particularly carp, and caninfect a range of mammals, including man.


Human infections may be present for manyyears without causing overt signs of clinical dis-ease. The development of clinical signs dependson the number of worms present, which can num-ber several thousands in heavy infections. Largenumbers of flukes in the bile ducts can cause cho-langitis (inflammation of the bile ducts) and gall-stones in man,115 and may possibly predispose tocholangiocarcinoma.143 The parasite may on occa-sions enter the pancreatic ducts, causing dilatation,fibrosis, and pancreatic stones.

Life Cycle

Adult flukes, which may live as long as 30years, inhabit the bile ducts of man and other fish-eating mammals. More than 100 species offish arereported hosts for the parasite in Asia. The lifecycle of adult fluke is shown in Figure 9-7.

Distribution

The parasite distribution of flukes is shown inTable 9-3. Infection rates have decreased in Ja-pan in recent years, but infections remain wide-

spread in China, Taiwan, and Korea. The preva-lence and distribution of infection is associatedwith the presence of susceptible snail and fishintermediate hosts and habits of the indigenouspopulations. It has been suggested that close to25 million people are infected in Asia, mostly insouthern China and North Korea.


Infection by flukes in man occurs by eatingraw, undercooked, or poorly preserved infectedfish. In some areas where the parasite occurs,fish is eaten raw in thin slices with rice (congee).The eating of raw freshwater fish has becomeincreasingly popular in several Far Easterncountries, and consequently the prevalence ofhuman infection is increasing. Infections outsideof Asia are usually imported. In most endemicareas, cats and dogs are infected in high num-bers. Levels of infection are maintained by thepractice of fertilizing ponds used to raise fishwith human excrement.

Diagnosis is based on the identification of thecharacteristic eggs in fecal samples, which haveto be differentiated from other trematode eggssuch as Heterophyes heterophyes, Metagonimusyokogawi, Haplorchis taichui, and Opisthorchisspecies (Table 9-^). Several serologic tests havebeen developed, but most are nonspecific. A re-ported enzyme-linked immunosorbent assay(ELISA) may be of value.30


In endemic areas, treatment of all infectedpersons and improved sanitation would helpcontrol infection by flukes. In areas where fishare raised in ponds, human and animal fecesshould be composted or sterilized before theyare applied as fertilizer to ponds.

Thorough cooking of all freshwater fish is themost effective means of controlling infection byflukes. Clonorchiosis can be prevented by avoid-ing eating raw, undercooked or improperly pick-led, salted, dried, smoked, or fermented fish inendemic areas. There is little or no informationon conditions affecting the survivability of theencysted infective stages in fish flesh. Educationprojects are of some benefit, but the eating ofraw or fermented freshwater fish has been a

Next Page

0C for at least four days or around -20 0C for atleast 12 hours. It is not known if cysterci can sur-vive meat fermentation conditions, but it wouldseem sensible to freeze or heat treat flesh ingre-dients wherever possible.

TREMATODES (FLUKES)

Trematodes or flukes are unsegmented,dorso-ventrally flattened, leaf- or lancet-shapedparasitic worms. The life cycles of the parasiticflukes discussed in the text are indirect, involv-ing a variety of intermediate hosts. A generaldescription of food-borne flukes and their classi-fication is provided in Table 9—1.

Clonorchiosis

Clonorchis sinensis (Opisthorchis sinensis), theChinese liver fluke, is a trematode infection that isfound throughout southeast Asia. The parasite isacquired by eating raw or poorly cooked or pre-served freshwater fish, particularly carp, and caninfect a range of mammals, including man.


Human infections may be present for manyyears without causing overt signs of clinical dis-ease. The development of clinical signs dependson the number of worms present, which can num-ber several thousands in heavy infections. Largenumbers of flukes in the bile ducts can cause cho-langitis (inflammation of the bile ducts) and gall-stones in man,115 and may possibly predispose tocholangiocarcinoma.143 The parasite may on occa-sions enter the pancreatic ducts, causing dilatation,fibrosis, and pancreatic stones.

Life Cycle

Adult flukes, which may live as long as 30years, inhabit the bile ducts of man and other fish-eating mammals. More than 100 species offish arereported hosts for the parasite in Asia. The lifecycle of adult fluke is shown in Figure 9-7.

Distribution

The parasite distribution of flukes is shown inTable 9-3. Infection rates have decreased in Ja-pan in recent years, but infections remain wide-

spread in China, Taiwan, and Korea. The preva-lence and distribution of infection is associatedwith the presence of susceptible snail and fishintermediate hosts and habits of the indigenouspopulations. It has been suggested that close to25 million people are infected in Asia, mostly insouthern China and North Korea.


Infection by flukes in man occurs by eatingraw, undercooked, or poorly preserved infectedfish. In some areas where the parasite occurs,fish is eaten raw in thin slices with rice (congee).The eating of raw freshwater fish has becomeincreasingly popular in several Far Easterncountries, and consequently the prevalence ofhuman infection is increasing. Infections outsideof Asia are usually imported. In most endemicareas, cats and dogs are infected in high num-bers. Levels of infection are maintained by thepractice of fertilizing ponds used to raise fishwith human excrement.

Diagnosis is based on the identification of thecharacteristic eggs in fecal samples, which haveto be differentiated from other trematode eggssuch as Heterophyes heterophyes, Metagonimusyokogawi, Haplorchis taichui, and Opisthorchisspecies (Table 9-^). Several serologic tests havebeen developed, but most are nonspecific. A re-ported enzyme-linked immunosorbent assay(ELISA) may be of value.30


In endemic areas, treatment of all infectedpersons and improved sanitation would helpcontrol infection by flukes. In areas where fishare raised in ponds, human and animal fecesshould be composted or sterilized before theyare applied as fertilizer to ponds.

Thorough cooking of all freshwater fish is themost effective means of controlling infection byflukes. Clonorchiosis can be prevented by avoid-ing eating raw, undercooked or improperly pick-led, salted, dried, smoked, or fermented fish inendemic areas. There is little or no informationon conditions affecting the survivability of theencysted infective stages in fish flesh. Educationprojects are of some benefit, but the eating ofraw or fermented freshwater fish has been a

Previous Page

Snail(first intermediate host)

Figure 9-7 Life cycle of Clonorchis sinensis. Source: © Crown Copyright.

practice among Asian communities for centu-ries, and many cultures would be reluctant tochange their habits. Freezing fish for a week at-10 0C would be beneficial, but even frozen fishhas been incriminated with Clonorchis out-breaks in nonendemic areas.

Echinostomosis

Trematodes of the family Echinostomatidaeare intestinal parasites of birds and mammalswith approximately 16 species reported fromhumans. Infections occur mainly in southeast

Hatched miracidiumin water

Free-swimmingcercaria

Fish(second intermediate host)

Clonorchis sinensis Egg in feces

Adult worm in liver

encysted metacercariae

Human infection throughingestion of raw or poorlypreserved fish containingencysted metacercariae

Table 9-3 Food-Borne Helminths (Trematodes)


Fish

Molluscs (snails, clams),fish

Salads (lettuce,watercress), alfalfa juice

Aquatic plants (watercaltrop, water chestnut)

Fish

Fish

Fish

Fish

Crustaceans, crabs,crayfish

Main Hosts (Final)

Man, dogs, cats, pigs, rats,mink

Wide variety of birds andmammals, man

Cattle, sheep, goats, pigs,horses, camelids, rabbits,wild herbivores, man

Cattle, sheep, goats,zebras, man

Man, pigs

Dogs, cats, fish-eating birdsand mammals, man

Dogs, fish-eating mammals(foxes, racoons, mink)

Dogs, cats, fish-eatingmammals, man

Cats and fish-eatingmammals

Man, dogs, cats, fish-eatingmammals (otters,mongooses, tigers,leopards, wolves,mustelids), rats, pigs,monkeys

Distribution

China, Korea, Japan, Taiwan, Vietnam

Far East (Phillipines, Thailand, Indonesia,Taiwan, Korea), Africa

Worldwide

Africa, parts of Asia, Hawaii

Asia (China, Taiwan, Thailand, Vietnam,Laos), India, Indonesia

Middle East (Egypt, Turkey), Balkans,southern Europe (Spain), and parts ofAsia (China, Japan, Korea)

North Amercia, Siberia

Southeast Asia (Thailand, Laos) Russia,Siberia, Kazakstan, parts of Europe(Poland, Germany) Southeast Asia(Korea, Japan, Phillipines)

Asia (China, Korea, Japan, Taiwan,Phillipines, Vietnam, India)

Central Africa

Central and South America (Peru,Ecuador, Colombia, Honduras, ElSalvador, Mexico)

Parasite

Clonorchis (Opisthorcis)sinensis

Echinostoma spp.*

Fasciola hepatica

Fasciola gigantica

Fasciolopsis buski

Heterophyesheterophyes

Nanophyetes salmincola

Opisthorcis viverrini

Opisthorcis felineus

Paragonimuswestermani

Paragonimus africanus

Paragonimus mexicanus

Parasite Class

Trematoda(flukes)

*Approximately 16 species reported in man.Source: © Crown Copyright.

Asia and involve species of the generaEchinostoma and Hypoderaeum.


Infections with echinostomids are usually ofshort duration and cause few problems unlesslarge numbers of fluke are present. Inflamma-tory lesions with shallow ulcers in the mucosamay develop at the site of the fluke attachment.Heavy infections may cause clinical symptomsof abdominal pain, flatulence, diarrhea, anor-exia, and edema. Diagnosis is based on findingthe eggs in the feces that must be differentiatedfrom those produced by Fasciola hepatica orFasciolopsis buski (see Table 9-4).

Life Cycle

These small flukes are found in the small in-testine of birds and mammals. The life cycle in-volves both snails and a range of second inter-mediate hosts, including other species ofmolluscs, fish, or tadpoles. Man acquires infec-tions after eating the second intermediate hostraw or partially cooked.

Distribution

The distribution of human echinostomid in-fections is provided in Table 9-3. The highestprevalence rates occur in the Philippines, Thai-land, Indonesia, Taiwan, Korea, and Africa.25


Infections are acquired by eating raw snails,clams, raw or undercooked frogs and tadpoles,or raw or insufficiently cooked fish. Manypeople in endemic areas are subclinically in-fected. Tourists may become infected when vis-iting endemic areas and sampling local cuisine.An outbreak of gastroenteritis caused byechinostome infection was reported in a group ofAmericans returning from a tour in Africa.129


Ensuring that snails, clams, tadpoles, frogs,and fish are well cooked before eating can pre-vent echinostomid infections. As with otherfood-borne trematode infections, there is a pau-

city of information concerning conditions af-fecting survivability of infective stages in food.

Fasciolosis

F. hepatica and F. gigantica (liver flukes) aretrematodes that live in the bile ducts of domesticand wild herbivores and occasionally infectman. In animals, fascioliosis occurs predomi-nantly in herbivores, most commonly affectingsheep, followed by cattle.


Human infection with F. hepatica and F.gigantica is often mild or subclinical, with theseverity of the infection dependent on parasitenumbers and the duration of infection. Seriousinfections, with a large number of parasites, mayproduce biliary stasis, liver cirrhosis, and chole-cystitis and gallstones, occasionally leading tojaundice.

Diagnosis in humans is based mainly on his-tory and clinical signs and must be differentiatedfrom other hepatic infections. Fecal examinationis not appropriate because the flukes rarely de-velop to maturity.

Life Cycle

Adult flukes are found in the bile ducts of theliver. The life cycles of F. hepatica and F.gigantica are similar but involve different snailintermediate hosts of the genus Lymnaea. Theprepatent period is approximately 12 weeks, andadult fluke can live in the bile ducts for severalyears.

Distribution

F. hepatica is distributed throughout theworld; F. gigantica occurs in Africa, severalAsian countries, and Hawaii. Human F. hepaticainfections occur sporadically, or in outbreaks,and have been recorded in numerous countries inthe Americas, Europe, Africa, and Asia. Themost extensive epidemics have occurred inFrance, with infection due to eating watercressthat was contaminated by metacercariae.98 Thedistributions of both parasites are summarized inTable 9-3. Fascioliosis is a common disease of

Differentiate from

Trichuris (protuding polar plugs)

Hymenolepis

Clonorchis/Heterophytes/MetagonimusOpisthorcis/ClonorchisHeterophytes/MetagonimusClonorchis/Opisthorcis

FasciolopsisFasciola

Parasite

Capillaria

Taenia soliumTaenia saginata

OpisthorcisClonorchis

HeterophytesMetagonimus

Paragonimus

Echinostoma

Fasciola hepaticaFasciolopsis buski

Size (\im)

35-45 x 21

35-40 x 30-35

25-35x10-20

25-30x15-17

80-120x48-60

80-120x60-90

130-145x70-90

Table 9-4 Worm Eggs in Human Feces

Description

Small to medium-sized eggs (< 75 |im)

With weakly protuding polar plugs

Embryonated

-spherical with hexacanth embryo

-with conspicuous operculum

-with inconspicuous operculum

Large eggs (> 75 jim)

Unembryonated

-with operculum


sheep, goats, and cattle in many parts of theworld. Economic losses due to Fasciola infec-tions are the result of liver condemnation,stunted growth, and reduced milk, meat, andwool production.


Man is infected with fascioliosis mainlythrough consuming salads of watercress (Nastur-tium officinale) containing metacercariae. InFrance, where watercress salad is popular, humaninfection is more frequent than in other Europeancountries. Contaminated lettuce or other plantsthat are eaten raw can sometimes serve as a sourceof infection, as can water from irrigation canals orother receptacles. Alfalfa juice has also been im-plicated in places where it is a customary drink.


Human infection can be reduced by avoidingeating watercress of wild or unknown origin andby ensuring that watercress is cultivated undercontrolled conditions that exclude animals or thepossibility of snail infestation. Treating live-stock limits levels of contamination, and controlof the intermediate snail host can be attemptedthrough the use of molluscides or effective landdrainage to reduce available snail habitats.

Fasciolopsiosis

Fasciolopsiosis is a disease that is caused bythe largest intestinal fluke, F. buski. This para-site infects man through the ingestion of raw,contaminated water plants.


Most infections are asymptomatic, but heavyinfections may produce diarrhea, abdominalpain, and occasionally intestinal obstructionleading to malnutrition and death.12'106 Diagnosisis confirmed by identifying the eggs in feces thathave to be differentiated from eggs of Fasciolaspp. (Table 9-4).

Life Cycle

Adult F. buski flukes are found in the duode-num. The life cycle is similar to that of Fasciola,

but intermediate hosts are freshwater snails ofthe genera Planorbis and Segmentina. Man (andpigs) becomes infected by eating contaminatedaquatic plants.

Distribution

It was estimated that approximately 10 mil-lion people were infested with fasciolopsiosisworldwide in the late 1940s.12 Current estimateson prevalence are not available, but evidencesuggests that the endemic range is expand-ing,68'71 although it is still confined to southeastAsia. The distribution of F. buski is summarizedin Table 9-3.


Infection with fasciolopsiosis is transmittedby the ingestion of raw contaminated aquaticplants such as water caltrop, water chestnut, wa-ter hyacinth, and water bamboo. Pigs are the res-ervoir host, but in Muslim countries, humansmay be the main reservoir of infection.


The disease is easily preventable by avoidingraw or uncooked aquatic plants in endemic ar-eas. The introduction of good sanitation facili-ties limits contamination of local water coursesand ponds.

Heterophyosis

More than 10 species of intestinal flukes ofthe family Heterophyidae have been reported inman and fish-eating mammals, the most impor-tant of which are Heterophyes heterophyes andMetagonimus yokogawai. These flukes are ac-quired by eating raw or poorly cooked or pre-served freshwater fish.


Light infections of intestinal flukes are usu-ally asymptomatic, but heavier infections causechronic, intermittent, often bloody, diarrhea; up-per abdominal pain; anorexia; nausea; vomiting;and weight loss. Eggs or worms entering theCNS may cause clinical symptoms that are simi-lar to those for cerebral hemorrhage or epilepsy.

In the heart, there may be signs of myocarditis orchronic congestive heart failure leading to death.

Diagnosis is based on finding and identifyingcharacteristic eggs in the feces. The eggs ofthese flukes are very similar, and specific identi-fication is difficult. They also have to be differ-entiated from those of C. sin en sis andOpisthorcis (Table 9-4). Serological tests arenot available.

Life Cycle

These small flukes live in the small intestines ofthe final hosts. The life cycles are similar to otherflukes involving both snail and fish intermediatehosts (e.g., mullet, tilapia, trout, and salmon).

Distribution

The distributions of H. heterophyes and Myokogawai are summarized in Table 9-3.


Intestinal fluke infections are acquired by eat-ing raw, undercooked, or poorly preservedfreshwater or marine fish and shrimps in en-demic areas. Endemnicity is maintained by con-tamination of water by human or animal feces.


The prevention of human infection relies onavoiding eating raw or improperly cooked, salted,or pickled infected freshwater fish in endemic ar-eas. Again, there is little or no information con-cerning conditions affecting survivability of infec-tive stages in fish or crustaceans. Education andgood sanitation would be helpful for control. Rawfish should not be fed to dogs or cats.

Nanophyetiosis

The intestinal trematode, Nanophyetessalmincola (Troglotrema salmincold), is foundin dogs and fish-eating mammals of eastern Si-beria and North America. It transmits a rickett-sia, Neorickettsia helminthoeca, the cause of"salmon poisoning" in these animals. Infectionwith this parasite has been reported in humans inNorth America.51


Mild fluke infections in man are asymptom-atic. Heavy infections (> 500 fluke) cause fre-quent bowel movements or diarrhea, abdominaldiscomfort, nausea, vomiting, weight loss, andfatigue. Human cases of salmon poisoning dueto the rickettsia, N. helminthoeca, have not beenreported. Diagnosis is based on the identificationof eggs in the feces or recovery of the adultworms following treatment.

Life Cycle

The small adult flukes live in the intestinal wallof the host. The life cycle involves both snail andfish intermediate hosts. Infection in man is ac-quired by eating uncooked infected fish (salmonand trout) containing encysted metacercariae.

Distribution

The parasite has been reported from the Pa-cific coast of North America and eastern Siberia.Its prevalence is unknown.


Animals and man acquire the infection by eat-ing raw or uncooked fish, although there is onereport of the infection being acquired by han-dling salmon.73


In endemic areas, the main preventative mea-sure is education in an effort to change the habitof eating raw or undercooked fish. There is noinformation on survivability of metacercariae infish-based fermented foods. Dogs and catsshould not be fed raw fish and should be keptaway from salmon rivers and streams.

Opisthorchiosis

Two species of Opisthorchis infect humansand cause a disease similar to clonorchiosis: O.viverrini is found in southeast Asia, and O.felineus (the cat liver fluke) is found in parts ofEurope.


Infection with Opisthorchis flukes has a sig-nificant public health importance because theirpresence often causes serious disease and islinked with cholangiocarcinoma and hepaticcarcinoma. Most infections are asymptomatic,depending on the level and duration of infection.Adult fluke in the bile ducts, gallbladder, andoccasionally the pancreatic duct cause thicken-ing of the ducts, predisposing to cholangio-carcinoma and hepatocellular carcinoma.74 Di-agnosis is made by identifying eggs in the feces,although eggs cannot be differentiated fromthose of C. sinensis and other fluke speciespresent in the intestines. An antibody ELISA59

and a molecular DNA method146 have been re-ported for diagnosis.

Life Cycle

Opisthorchis species are similar morphologi-cally and in life cycle to Clonorchis sinensis.Adult fluke can live as long as 20 years.

Distribution

The distributions of O. viverrini and O.feline-us are shown in Table 9-3. It is estimated thatmore than 7 million people are infected with O.viverrini and 4 million with O. felineus. Preva-lence rates in highly endemic areas are greaterthan 90% for O. viverrini and 85% for O. fe-lineus. In some parts of Europe, more than 85%of cats are infected with O. felineus.


Man and other definitive hosts become in-fected with Opisthorchis flukes by eating raw orundercooked fish. In Europe, metacercariae arecommon in freshwater fish such as orfe, bream,tench, and barbel. In parts of Asia, fish are eatenraw, especially in the rainy season when the fishare most abundant. In Thailand, fish are raised inponds and the ponds are fertilized with humanand animal feces. In areas endemic for O.felineus, the freshwater fish is eaten raw, dried,salted, or fermented. Cats and dogs in southeastAsia are important reservoirs of O. viverrini.


The main preventative measure is througheducation of the population to avoid eating rawor poorly preserved fish. Food habits are tradi-tional and are often difficult to break. Saltingfish in 5-15% saline for three to ten days de-stroys metacercariae, as does freezing at -10 0Cfor five days. It is not known if fermentationconditions kill metacercariae. Changes in sanita-tion may limit contamination of fish ponds.Snail control and eradication of reservoir hostsare not feasible. Mass anthelmintic treatmentwith praziquantel targeted at heavily infected in-dividuals as part of community-based controlprograms may be of benefit, although re-infec-tion after treatment is common.160

Paragonimosis

Approximately 40 species of Paragonimushave been described from around the world, ofwhich only 8 species are considered to be impor-tant. P. westermani is widespread in the Far Eastand is responsible for most cases of pulmonaryparagonimosis. Other species that infect man in-clude P. heterotremas, P. pulmonalis, P. skry-jabini, P. miyazakii in Asia, P. uterobilateralisand P. africanus in Africa, and P. mexicanus inLatin America. P. kellicotti is a North Americanspecies reported from animals and only rarely inhumans.


Pulmonary infections with Paragonimus areusually asymptomatic except for occasional he-moptysis. The pathology associated with infec-tions of Paragonimus varies depending on thelocation and numbers of flukes present. Adultflukes in the lung parenchyma may cause gen-eral malaise, coughing, and dyspnoea due to in-flammation and cyst formation around the fluke.Extrapulmonary infections may produce a cuta-neous larval migrans and abscess formation inthe skin and viscera. Brain and spinal cord in-volvement may lead to headaches, seizures,paraplegia, and occasional deaths. Examination

of sputum or feces will usually reveal the pres-ence of the typical brown, thick-shelled, opercu-late eggs (Table 9^4).

Life Cycle

Adult flukes are usually located in pairs ingranulomatous capsules in the lungs. The lifecycle, which involves snail and crustacean inter-mediate hosts, is shown in Figure 9-8. Some in-fections are reported to persist for 10 or even 20years.

Distribution

The distributions of P. westermani,P. africanus, and P. mexicanus are summarizedin Table 9-3. Data on prevalence rates are lim-ited, but it has been estimated that 1-1.5 millionpeople are infected with P. westermani in Ko-rea.106 A study in Cameroon showed a preva-lence of 5.6% for P. africanus.9®


The source of infection for man and other de-finitive hosts is infected freshwater crustaceans.In man, this usually occurs following the inges-tion of infected raw or poorly cooked, pickled,or salted crabs or crayfish. In some areas ofChina, crabs are soaked in rice wine for a periodof time before eating (i.e., drunken crabs). Inother parts of China, crayfish curd, raw crabsauce, or jam are eaten. In Thailand, raw fresh-water shrimp salad and crab sauce are eaten.Koreans eat uncooked crabs immersed in soysauce. Crabs are eaten roasted or raw in the Phil-ippines, and crab juice is used in preparing somemeat dishes. Infections are also possible by eat-ing animal meat containing immature flukes.Human infection has been reported in Japan fol-lowing the ingestion of uncooked slices of wildboar meat.106 Contaminated water and cookingutensils are also a source of infection.

Domestic and wild animals act as a reservoirof infection for Paragonimus spp. In endemicareas, human infection rates are sufficient tomaintain the cycle of infection through contami-nation of water courses.


As with other food-borne infections,paragonimosis can be prevented through educa-tion to prevent the consumption of raw orundercooked crabs and crayfish. The cooking ofcrustaceans at 55 0C for five minutes will killmetacercariae. Fermentation processes may beinsufficient to kill metacercariae, although directevidence is lacking. Improved sanitation helpsreduce the infection of intermediate hosts.

PROTOZOA

Protozoa are a diverse group of microorgan-isms with simple or complex structures con-tained within a single cell and equally simple orintricate life cycles that vary considerably de-pending on the species. General characteristicsof the protozoa described in the text are providedin Table 9-5.

Cryptosporidiosis

Protozoa of the coccidial genus, Crypto-sporidium, are small, intracellular parasites thatoccur throughout the animal kingdom and havebeen reported in many species of mammals,birds, reptiles, and fish.57'156 There are at leastnine recognized species of Cryptosporidium, ofwhich the most important is C. parvum, whichinfects a wide range of mammals includingman.56 C. parvum has been reported in 79 spe-cies of mammals and is highly prevalent in rumi-nants, particularly in young calves, and appearsto be age related.57'153 It is common in younglivestock, especially cattle and sheep, althoughpigs, goats, horses, and deer can be infected.Dogs, cats, and other pets are occasionally in-fected, but they do not appear to be an importantsource of infection to other hosts.


Infections with C. parvum are a significantcause of diarrhea, abdominal pain, nausea, vom-iting, and fever in humans.159 Diarrhea is moreprofuse in the immunocompromised, such asthose with AIDS; in these patients, infections are

Figure 9-8 Life cycle of Paragonimus westermani. Source: © Crown Copyright.

Development in first (snail)intermediate host

Free-swimmingcercariae

Hatched miracidiumin water

Metacercariae in second(crustacean) intermediate hosts

Paragonimus westermani

Egg in sputumorfeces

Adult wormin lung

Table 9-5 Food-Borne Protozoa — Classification

Phylum

Apicomplexa

SarcomastigophoraSubphylum Mastigophora

Subphylum Sarcodina

Subphylum Blastocysta+

Ciliophora

Genera*

CryptosporidiumCyclosporaSarcocystlsToxoplasma

Giardia

Entamoeba

Blastocystis

Balantidium

General Characteristics

Apical complex (visible with electron micro-scope) generally consisting of polar ring,rhoptries, microneme, conoid; all speciesparasitic

One or more flagella usually present introphozoites; asexual reproduction longitudinal

Pseudopoda or locomotive protoplasmic flowwithout discrete pseudopodia; asexualreproduction by fission

Pleomorphic, brightly refractile cysts with acentral vacuole; amoeboid form may occur

Cilated organisms with micronucleus andmacronucleus

*Genera of parasitic protozoa referred to in text onlytJiang and He (1993) suggest that Blastocystis should be classified in its own subphylum within the phylum sarcomastigophora81


of long duration, often followed by death. In theimmunocompetent, infections are self limitingand last 7-10 days.

Life Cycle

The life cycle of infection with C. parvum istypically coccidian, with all development stagesoccurring within the same host. The infectivestage is the oocyst, which is passed into the envi-ronment by infective hosts in the feces. For moredetailed descriptions of the life cycle, see Tay-lor.154

Distribution

Human infection with C. parvum has world-wide distribution and has become a commoncause of gastrointestinal infection (Table 9-6).In more countries of Europe and North America,the incidence of clinical infection is generallybetween 1-3%; elsewhere, rates of between4.9% in Asia to 10.4% in Africa have been sug-gested.159 The higher prevalence in less devel-

oped countries may relate to lack of clean water,sanitation, or general crowding. Seroprevalencestudies suggest a higher rate of exposure tocryptosporidiosis, with levels between 25% and35% in Europe and North America, suggestingthat many infections are asymptomatic.159


Transmission from host to host is via the oo-cyst stage, usually through fecal contaminationof water, food, or animal-to-person or person-to-person contact. Many mammals have beenfound to be naturally infected with C. parvumand have the potential, therefore, to act as reser-voir hosts. Person-to-person transmission is con-sidered to be the main route of transmission,147

although zoonotic transmission has been re-ported from calves and lambs, particularly fol-lowing educational visits to farms.148 Recent re-search suggests that two distinct strains of C.parvum (type 1, human, and type 2, animal) ex-ist, with differing epidemiological cycles.7'102

Table 9-6 Food-Borne Protozoa


Animal products (meat, milk, offal)unpasteurized cider, direct fecalcontamination, water

Fruit (raspberries), water, salad

Meat (beef)

Meat (pork)

Meat (lamb, mutton, pork), water,direct fecal transmission (cat feces)

Contaminated salad, fruit, processedfood, water, direct fecalcontamination

Contaminated food and water, directfecal contamination



Main Hosts (Final)

Man and mammals (cattle,sheep, pigs, goats,horses, dogs, cats)

Man, primates ^insecti-vorous mammals, snakes)

Man

Man

Cat, wild felids

Man, cattle, sheep, goats,horses, dogs, cats

Pigs, cattle, dogs, cats, man

Man, pigs, monkeys

Pigs, dog, rat, man

Distribution

Worldwide

Worldwide? Americas, Europe,India, parts of Africa andSoutheast Asia

Worldwide but low prevalence(Europe— France, Germany,Poland, Czech Republic,Slovakia)

Eastern Euro^

Worldwide

Worldwide

Worldwide

Worldwide?

Worldwide

Parasite

Cryptosporidium parvum

Cyclospora cayetanensis

Sarcocystis bovihominis

Sarcocystis suihominis

Toxoplasma gondii

Giardia intestinalis

Entamoeba histolytica

Blastocystis hominis

Balantidium coll

Parasite Class

Apicomplexa(Coccidia)

Mastigophora(Flagellates)

Sarcodina(Amoebae)

Blastocysta

Ciliophora(Ciliates)


Food-borne transmission of cryptosporidiosishas been reported following the consumption ofcertain foods, notably raw sausage, offal, chickensalad, and milk.15'26"28'65'67'83'112'157'175 An outbreakof cryptosporidiosis has been reported amongpeople drinking freshly pressed, unpasteurized ci-der, possibly resulting from fecal contamination offallen apples by animals.107'121 Outbreaks ofcryptosporidiosis associated with contaminateddrinking water have been reported from a numberof countries, including large outbreaks in theUnited States and the United Kingdom.17


Contaminated water is a major source of in-fection and appears relatively common. It canresult in large-scale outbreaks of disease. Im-provement in water catchment may reduce thepotential health risks associated with the pres-ence of oocysts in water supplies, but wouldprobably not eliminate the problem. The safetyof bottled water varies from supplier to supplier.Heating water to temperatures greater than 72.40C for one minute, or greater than 64.2 0C for twominutes of a five-minute heating cycle inactivatesC. parvum oocysts.54 Where ooysts ofCryptosporidium are detected in domestic watersupplies, boil notices may be issued with the rec-ommendation to boil water before drinking.17

Concentration of oocysts from water is possible byfiltering large volumes of water using cartridge ormembrane filtration, or smaller volumes using acalcium carbonate flocculation technique.64

The potential risks from food-borne crypto-sporidiosis come from the ingestion of fresh,raw, or uncooked foods. Effective means of con-trol include the avoidance of contamination ofvegetables and fruits in the field, strict hygienemeasures involving handling of food, and thesusceptibility of the organism to freezing andcooking. Evidence suggests that untreated milk,undercooked sausage meat, and offal have beenimplicated in outbreaks, but heat treatments suchas those used for the pasteurization of foods arelikely to destroy oocysts.9 Oocyst infectivity isalso destroyed by freezing or by freeze drying.Dried or frozen foods are therefore unlikely tobe sources of food-borne cryptosporidiosis. The

main risks would appear to be from fresh or re-frigerated food. The effects of fermentation pro-cesses and other nonheat processes on oocystsurvivability are not known.42 Concentrationtechniques have been described for the detectionof oocysts in food.91

Cyclosporiosis

Cyclospora are coccidian parasites reportedfrom a number of animals that have assumed im-portance more recently in man, with worldwidereports of the presence of "cyanobacterium-likebodies" in the feces of immunocompetent andimmunocompromised patients with diarrhea.117

These organisms are now recognized as Cyclo-spora cayetanensis and are a cause of protracteddiarrhea in humans.174 It is thought that initialcases of infection in man were either reported asunidentified coccidian parasites6 probably mis-identified as Isospora136 or Cryptosporidiumm ordescribed as cyanobacterium-like organisms.95

Parasites of this genus have been described in in-sectivorous mammals,50'61'62 nonhuman primates,and snakes.126


Infections with C. cayetanensis in man cancause protracted diarrhea with nausea, vomiting,anorexia, bloating, abdominal cramps, fatigue,and malaise.6'149 Among immunocompromisedhosts, Cyclospora infection has occurred pre-dominantly in HIV-infected individuals. Clini-cal illness in these patients is prolonged and se-vere,77 and is associated with a high rate ofrecurrence.123'145

Life Cycle

The life cycle of C. cayetanensis in animals istypically coccidian and is likely to be similar inhumans.154 Oocysts may require from two weeksto four to six months to become infective, de-pending on temperature.150

Distribution

Although Cyclospora appears to be widelydistributed throughout the world, the prevalenceof infection in man and animals worldwide is

unknown. Distribution is summarized in Table9-6. In Canada and the United States, approxi-mately 1 in 1,000 human fecal samples from di-arrheic patients were positive,18'116 although theprevalence in the general population was consid-ered to be generally lower. Prevalence rates of11% have been reported among allochthonousadults and children in Nepal during the rainyseason.77 Similar levels of infection have beenreported in HIV-positive adults in Haiti.123


Waterborne outbreaks of cyclosporiosis havebeen suspected,150 and drinking untreated waterduring foreign travel has been incriminated as afactor in the epidemiology of the disease.77 Morerecently, there have been several reports of food-borne transmission associated with eating fruitin North America. A large outbreak ofcyclosporiosis in the United States was associ-ated with the consumption of imported raspber-ries.75 Meselun (a mixture of various types ofbaby leaves of lettuce) was suspected of beingthe vehicle for one or possibly two outbreaks inthe United States in 1997.161 The consumption ofthe herb, basil, was suspected as as a source ofinfection at a luncheon in Virginia.120 Commer-cial freezing may kill the oocysts but has yet tobe proven.79 Although there have been no re-ports of cyclosporidiosis transmission in fer-mented foods, the distinct possibility exists, andlittle if any studies have been conducted onthe survivability of oocysts under fermentationconditions.


Precautions similar to those recommended forcryptosporidiosis should be taken wherever pos-sible. In Third World countries with poor sanita-tion, either water should be boiled or onlybottled water should be consumed. As most re-ports of cyclosporiosis occur following the con-sumption of fresh fruit and salads, these shouldbe washed thoroughly and food handlers shouldensure good personal hygiene. The detection ofCydospora on food, particularly fruit, is diffi-cult, even though methods for detecting proto-zoan contamination exist.

Sarcocystosis

Sarcocystis is one of the most prevalent para-sites of livestock infecting mammals, includingman, birds, and lower vertebrates.48 Approxi-mately 130 different species have been reported.The parasites derive their name from the intra-muscular cyst stage (sarcocyst) of the life cycle,which are found in virtually all skeletal musclesof the body, including the tongue, esophagus,diaphragm, cardiac muscle, and brain of the in-termediate host. Sarcocystis is found widelythroughout the animal kingdom. In domesticanimals, for example, there are three species ofSarcocystis in cattle (one species of which, S.hominis [S. bovihominis], is infective for man);four species in sheep (S. tenella and S.arieticanis [S. ovicanis] infect dogs andS. gigantea and S. medusiformis [S. ovifelis] in-fect cats); and three species in pigs, one ofwhich, S. suihominis, is infective for man.154

Food-borne transmission occurs through the in-gestion of raw or undercooked meat containingmature sarcocysts.


The ingestion of raw beef containing S.hominis can produce clinical signs of nausea,stomachache, and diarrhea within three to sixhours.5'137 The ingestion of infected pork con-taining S. suihominis is much more dramatic,producing clinical signs of bloat, nausea, loss ofappetite, stomachache, vomiting, diarrhea, diffi-culty breathing, and rapid pulse within 6-42hours. Sarcocystis may be responsible for sev-eral idiopathic diseases in man, including car-diac diseases such as cardiomyopathy8'70 andrheumatic diseases.70 It has also been suggestedthat Sarcocystis may be associated with muscleaches and fatigue as part of the chronic fatiguesyndrome (CFS).128

Life Cycle

The life cycle of Sarcocystis involves an inter-mediate herbivorous host and a final carnivoroushost.154 The final host becomes infected by in-gesting cysts contained in the muscle or nervoustissue of an intermediate host.

Distribution

There are two known species of Sarcocystis thatinfect man — S. hominis (S. bovihominis) andS. suihominis (Table 9-6). There is evidence tosuggest that oocysts of these species were previ-ously referred to as "Isospora hominis.'"1 Humansalso serve as the accidental intermediate host forseveral unidentified species of Sarcocystis.11^Levels of infection in man may reflect cookinghabits and preferences for undercooked meat.

Examination of various tissues suggests thatmost cattle worldwide are infected with Sarco-cystis^ with S. cruzi (S. bovicanis), which is in-fective to dogs, being the most prevalent. Infec-tion of cattle with S. (bovi)hominis appears to befairly low in many countries, with the possibleexception of Germany, where up to 63% ofcattle have been shown to be infected.142 Almostall adult sheep can be infected with Sarcocystis,but none are infective for man.48 The overallprevalence of Sarcocystis in pigs is low, be-tween 3-36% worldwide.48 S. suihominis hasbeen reported in European countries, with thehighest prevalence in Germany.16


In most cases, humans acquire gastrointestinalsarcocystosis by ingesting raw or undercookedmeat from cattle or pigs harboring mature S.(bovi)hominis or 5. suihominis cysts. Macroscopiccysts of Sarcocystis spp. are often removed atmeat inspection, although these cysts are usuallyof no significance to man. Some meat animalssuch as camels, llamas, water buffalo, and yaksharbor Sarcocystis in which the final host is un-known.48 In animals, chronic sarcocystosis cancause economic losses due to reduced qualityand quantity of meat. Additional economiclosses occur with those species that form macro-scopic cysts resulting in condemnation of wholecarcasses or affected parts after slaughter.


Although a widespread and common infectionin domestic animals, transmission to man throughthe ingestion of infected meat is a relatively rare

occurrence. Species of Sarcocystis are highlyhost-specific, and only two species are known toinfect man. The greatest potential risk thereforecomes from ingesting raw or uncooked beef orpork. Ensuring that meat is cooked thoroughly orfrozen can prevent infection. It is not known ifmeat fermentation processes will kill thesarcocysts that are present in meat.

To interrupt the life cycle and prevent infec-tion of livestock, carcasses, meat scraps, offal, orother raw or undercooked tissues should not beavailable to domestic or wild carnivores in areaswhere livestock are raised. The identification ofSarcocystis in meat can be made by direct obser-vation or by a meat digestion method and exami-nation for the presence of bradyzoites.48

Toxoplasmosis

Toxoplasma gondii is an obligate intracellularcoccidian parasite with a two-host life cycle inwhich oocyst production occurs in domestic andwild cats.53 The parasite differs from other coc-cidia, however, in that it shows a complete lackof specificity for hosts and tissues during itsasexual phase and can infect a wide range of ani-mals, including man. The parasite has world-wide distribution and is of both medical and vet-erinary importance.


Infections with T. gondii are usually asymp-tomatic, but can cause mild symptoms of lym-phadenopathy through to signs of anemia, en-teritis, fever, hepatitis, lymphadenitis,myocarditis, and pneumonitis. Symptoms areworse in immunocompromised individuals, inwhom infections can be severe, with cerebral in-volvement (encephalitis). The parasite is a sig-nificant cause of congenital infection and abor-tion in women.

Life Cycle

The life cycle of T. gondii is similar to Sarco-cystis in that it involves a carnivore (cat) finalhost but differs in lack of specificity for interme-diate hosts that includes virtually every speciesof warm-blooded animals and birds.

Distribution

Toxoplasma is widespread among man andanimals (Table 9-6) and is possibly the mostwidespread and prevalent protozoan parasite onearth.53 It has been estimated that more than 500million people are infected with Toxoplasma.45

In the United States and the United Kingdom,estimated infection rates in man are between20% and 40%, but are much higher in Europeancountries, with rates ranging from 50% to 80%.44

Infection appears to be more common in warmclimates and low-lying areas, probably relatingto climatic conditions favoring oocyst sporula-tion and survival.63'177 Encephalitis caused byToxoplasma is a major opportunistic infection inimmunocompromised individuals. It is esti-mated that 30% of AIDS patients seropositive toT. gondii will develop encephalitis.40 In Europe,congenital toxoplasmosis has been reported in1-6% of newborn children and is probably themost serious form of Toxoplasma infection.


Toxoplasmosis is acquired by ingestingsporulated oocysts or viable tissue cysts in meatfrom infected animals. The extent of human in-fection resulting from the ingestion of oocysts isnot known. The ingestion of oocysts has beenincriminated as the cause for the relatively highprevalence of Toxoplasma antibody in vegetar-ians in India.134 There are few reports of water-borne transmission of Toxoplasma.13 In contrast,food-borne transmission by the ingestion of tis-sue cysts in raw or undercooked meat from a va-riety of livestock and game animals has beenwell documented as a major source of human in-fection.45'53 Pigs, sheep, and goats commonlyharbor tissue cysts, whereas cattle appear to berelatively resistant to infection.53 The ingestionof rare or undercooked meat is a common causeof infection in many countries,44 with culturalhabits an important factor in the epidemiology.In France, for example, where raw orundercooked meat is regularly eaten, 84% ofFrench women have antibodies to T. gondii. Un-pasteurized milk from infected goats has alsobeen implicated as a source of infection.135'138

Occupational risks of infection may include cer-tain farming practices, and pregnant women andother people in special risk groups should avoidcontact with sheep at lambing time.23


Food-borne transmission of toxoplasmosisthrough the ingestion of raw or undercookedmeat or other animal produce such as milk iswell documented. Adequate cooking kills the or-ganism, as does pasteurization. Therefore, toavoid infection, all meat should be cooked thor-oughly before eating. Cooking at 60 0C or higherfor 3 !/2 minutes or longer renders Toxoplasmacysts noninfectious.47'55 This temperature is nec-essary at the core of the meat, not just the sur-face. Microwave ovens should not be used forinitial cooking because of the uneven distribu-tion of heat resulting in hot and cold spots.54

Freezing to -20 0C also kills some strains ofToxoplasma, but is less dependable than cook-ing. Salting, curing, and pickling also kills cystspresent in meat,171 but it is not known if meat fer-mentation processes render cysts nonviable. Theirradiation of meat at 50 kilorads or greater ren-ders Toxoplasma cysts noninfectious.46 The re-duction or elimination of oocyst contaminationfrom cat feces also helps prevent toxoplasmosis.Domestic cats should be prevented from huntingbirds and rodents and should be fed preferablydry or canned food. Litter trays should be used inthe home and emptied daily. Gardeners handlingsoil should wear gloves and wash their handsthoroughly before eating or handling food. Catson farms should be prevented from enteringfeed-storage facilities and animal quarters. Fi-nally, water from lakes, ponds, streams, or otheruntreated sources should be boiled before drink-ing or being used to wash food.

Giardiosis

Although many species of Giardia (Table 9—5) have been reported in the literature, it is gen-erally considered that there are just three struc-tural types or species. Giardia infections occurwidely in mammals, birds, reptiles, amphibians,and fish. The species affecting man and the ma-

jority of domestic animals is Giardia duodenalis(synonymous, G. intestinalis, G. lamblia, andLamblia intestinalis). Other morphologicallydifferent species are G. muris, which is identi-fied in rodents, birds, and reptiles, and G. agilis,from amphibians.


Giardia infections can cause nausea, inappe-tence, abdominal discomfort, fatigue, foul-smelling watery diarrhea, flatulence, and ab-dominal distention, lasting from only a fewdays to several months. Chronic infections leadto malabsorption, weight loss, and debilitation.Mortality has seldom been reported.

Life Cycle

Giardia are flagellate protozoans that are nor-mally found adherent to epithelial surfaces of thesmall intestine, especially the middle to lowerareas of the villi. The life cycle of Giardia issimple and direct, the trophozoite stage (motileforms) dividing by binary fission to produce fur-ther trophozoites. Intermittently, trophozoitesform resistant cyst stages that pass out in the fe-ces of the host.2

Distribution

Infection with G. intestinalis has been in-creasingly recognized and has been reported asthe most common gastrointestinal protozoanparasite in the world in man168 (Table 9-6). Veryhigh prevalence rates of between 20% and 60%are common, particularly in children in develop-ing countries 2 1 ,49,52,68,78, i oo, 1 03, 1 04

The existence of Giardia in domestic animalshas been known for many years, but during thattime, little information has become available onprevalence, pathogenicity, and the diseasecaused by these parasites in their hosts. Giardiahas been reported in cattle, sheep, goats, horses,dogs, cats, rodents, and psittacines, with varyingprevalence rates in these hosts.22'86'87'105'151'155


Transmission of the resistant cyst stage is viathe fecal-oral route through contaminated water,contaminated food, or person-to-person, animal-

to-animal contact. Cysts can remain infectious forseveral months under moist and cool conditions,but lose infectivity under dry and hot conditions.

Infection with G. intestinalis is generally as-sociated with conditions of poor hygiene andsanitation, including poor control of water qual-ity.133'168 Infection is common among childrenwho attend daycare centers, but is also associ-ated with travel.19'82'168

Reports of food-borne transmission of Giar-dia have been made only occasionally. Cystshave been detected in salad, fruit, and other fooditems.10 Outbreaks of giardiosis have occurredfollowing the consumption of salmon and creamcheese dip,119 noodle salad,127 sandwiches,170 andfruit salad.130

Waterborne outbreaks of giardiosis have beenreported in the United States and the UnitedKingdom.80 In the United States, Giardia hasbeen identified more often than any other patho-gen in waterborne outbreaks that result in ill-ness. Because Giardia is so widespread amongdomestic and wild animal hosts, it has the poten-tial for zoonotic transmission.


The control of giardiosis is generally based ongood personal hygiene, proper sanitation, andthe treatment of drinking water by a combinationof filtration and disinfection. Reports of food-borne infections with Giardia are relatively un-common. To prevent giardiosis, food handlersshould ensure good personal hygiene, and freshfruit and salads should be washed thoroughly.Fermented cheeses may be a source of giardiosisand, wherever possible, milk ingredients shouldbe pasteurized.

Amoebiosis

Entamoeba histolytica is one of six parasiticamoebae of the genus Entamoeba known to in-fect humans20 (Table 9-5). In man, the primaryroute of transmission is from human to humanby the fecal-oral route, usually associated withpoor hygiene or poor water quality. Infectionswith£. histolytica occur worldwide but are moreprevalent in the tropics. There are two morpho-

logically identical forms of E. histolytica — onenonpathogenic and the other pathogenic.20 E.histolytica has also been recorded in cattle, pigs,dogs, cats, and monkeys.94


Most human infections with Entamoeba areasymptomatic, but approximately 10% of in-fected individuals develop clinical symptoms ofdysentery with bloody or mucoid diarrhea last-ing from a few days to several months. Ulcer-ation of the colon may occur, which mayprogress to ulcerative lesions in the liver, skin,or brain.

Life Cycle

Trophozoites present in the intestine divide bybinary fission. Transmission occurs when the tro-phozoites round up, become smaller, and form acyst. Nonpathogenic forms of the organism nor-mally live in the lumen of the large intestine.Pathogenic forms invade the mucosa, causing ul-ceration and dysentery. From there, they may becarried via the portal system to the liver and otherorgans, where large abscesses may form.

Distribution

It has been estimated that approximately 480million people, or 12% of the world's popula-tion, are infected with E. histolytica, with an an-nual mortality of 40,000 to 110,000 people69

(Table 9-6). Approximately 10% of thosepeople who are infected every year have clinicalsymptoms.164 Five percent of the U.S. popula-tion were estimated to be infected, but the preva-lence of E. histolytica appeared to be declining.97

The incidence of E. histolytica in animals isunknown. It has been found in naturally infectedmonkeys of many species throughout the world.Infection in dogs has only been reported sporadi-cally and often through human contacts. Naturalinfections are apparently rare in pigs and have tobe differentiated from other species that occur inthe pig.


The ingestion of food and drink contaminatedwith E. histolytica cysts from human feces and

direct fecal-oral contact are the most commonmeans of infection. Transmission from human tohuman by the fecal-oral route is usually associ-ated with poor hygiene or poor water quality.Cysts can be introduced into the mouth by soiledhands of food handlers, family members, hospi-tal personnel, and other close personal contacts;by food contamination via flies; and by watercontaminated with sewage. The transmission ofE. histolytica by water is common in ThirdWorld countries where much of the drinkingwater is untreated. The use of human feces forfertilizer is also an important source of infec-tion.21 Recognized high-risk groups for infectioninclude travelers, immigrants, migrant workers,immunocompromised individuals, and male ho-mosexuals. Zoonotic infection from animals hasnot been recorded.


The main risk of infection with E. histolyticaappears to be through drinking untreated or con-taminated water or food. Good personal hy-giene, especially thorough washing of hands byfood handlers, good sanitation, and high-qualitywater treatment are essential for preventingtransmission.

Blastocystosis

Blastocystis was for many years described asa yeast but is now considered to be a protozoanin the subphylum, Blastocysta81 (Table 9-5).The organism is found in the intestinal tract ofman and in many animals, including monkeys,pigs, birds, rodents, snakes, and invertebrates.


Infection with Blastocystis hominis is gener-ally asymptomatic, but it is becoming increas-ingly associated with gastrointestinal disease inman.178 Symptoms include diarrhea, abdominaldiscomfort, anorexia, flatulence, and other non-specific gastrointestinal effects.

Life Cycle

Various life cycles have been proposed forBlastocystis. Transmission is thought to occur

either through the ingestions of a vacuolar or incyst form (e.g., seen in feces).154

Distribution

The prevalence of B. hominis in man is veryvariable (Table 9-6). Studies in man show vary-ing levels of infection in both healthy patientsand patients with diarrhea.101'113 Infection ap-pears to be more common in adults than in chil-dren.43'84'132

Blastocystis spp. have been reported in a num-ber of animals, but no prevalence figures areavailable. In pigs, cross-transmission studieshave indicated a low host specificity and uncer-tain pathogenicity.122


It is generally presumed that B. hominis istransmitted by the fecal-oral route in a similarmanner to other enteric protozoa.66 Waterborneand food-borne transmission are possible routesof infection66 but have not been investigated. In-fections in man have been linked with foreigntravel, although one study suggested an associa-tion with exposure to pets or farm animals.43


B. hominis is increasingly reported as an in-testinal infection of man, but it is not clearwhether it acts as a pathogen. Transmission is bythe fecal-oral route, and waterborne and food-borne transmission are potential routes of infec-tion, but have not as yet been investigated to anygreat extent. Prevention is as with other intesti-nal protozoa through good hygiene, especiallythorough washing of hands by food handlers andthe instigation of good food handling practices.

Balantidiosis

Balantidium coll occurs in man, other pri-mates, pigs, and occasionally in dogs, rats, andruminants94 (Table 9-5).

The primary host appears to be the pig, inwhich it is generally regarded as a commensal ofthe large intestine, where it lives on starch, in-gesta, and bacteria.94 The parasite has occasion-

ally been reported as a food-borne or waterbornezoonotic infection of man.


Infection with B. coli is usually asymptomaticbut may, on occasion, produce superficial todeep ulcers that result in intermittent diarrhea ordysentery. Rare complications include perfora-tion of the bowel wall and extra-intestinal infec-tion of the liver, vagina, ureter, and bladder.

Life Cycle

Trophozoites in the lumen or tissues multiplyby binary fission and, under certain conditions,transform into cysts that pass from the body inthe feces.

Distribution

The Balantidium organism is found world-wide (Table 9-6) in humans, but the prevalenceof disease based primarily on case reports ap-pears low. Surveys of fecal samples collectedthroughout New Guinea revealed prevalencerates of infection with Balantidium as high as29% in some areas in which people lived in inti-mate contact with pigs.


Transmission via the fecal-oral route is fromhuman to human and possibly from pigs or ratsto humans. Transmission is facilitated in con-fined environments in which hygiene is poor.The cyst form is the source of infection, andcysts can remain viable for days or weeks inmoist pig feces. Flies are possible mechanicalvectors. Although pigs have been implicated as amajor source of human infection, their role isstill a source of controversy. In a review of theliterature, more than 50% of the people with bal-antidiosis reported contact with pigs.166 In onecase, a family claimed that infection followedingestion of raw pork sausage. A waterborneoutbreak implicating pigs as the source of infec-tion has been reported.167

Balantidiosis in monkeys and primates is nor-mally an endemic infection maintained by theanimals themselves.


Infection in man is likely to occur throughcontamination of water or uncooked vegetables.Prevention is the same as for other protozoa.Under severe environmental conditions, poten-tially contaminated water should be boiled be-fore use. Vegetables should be washed thor-oughly before use.

CONCLUSION

Food-borne parasitic diseases are common inmany cultures, and human infections can arisefollowing the ingestion of a wide variety of food

REFERENCES

1. Ackers, J. P. (1997). Gut coccidia — Isospora,Cryptosporidium, Cyclospora and Sarcocystis. SemGastrointest Dis 8, 33^4.

2. Adam, R. D. (1991). The biology of Giardia spp.MicrobiolRev 55, 706-732.

3. Allan, J. C., Avila, G., Garcia Noval, J., Flisser, A. &Craig, P. S. (1990). Immunodiagnosis of taeniasis bycoproantigen detection. Parasitol 101, 473-477.

4. Anwar, M., Rak, H. & Gyorkos, T. W. (1979). The in-cidence of Gongylonema pulchrum from cattle inTehran, Iran. Vet Parasitol 5, 271-274.

5. Aryeetey, M. E. & Piekarski, G. (1976). SerologischeSarcocystis-Studien an Menschen und Ratten. Z.Parasitenkd 50, 109-124.

6. Ashford, R. W. (1979). Occurrence of an undescribedcoccidian in Paua New Guinea. Ann Tr op MedParasitol 73, 497-500.

7. Awad-el-Kariem, F. M., Robinson, H. A., Pety, F.,McDonald, V., Evans, D. & Casemore, D. (1998). Dif-ferentiation between human and animal isolates ofCryptosporidium parvum using molecular and biologi-cal markers. Parasitol Res 84, 197-201.

8. Azab, M. E. & El-Shennaway, S. F. A. (1992). Investi-gation of Sarcocystis as a causative agent in cardiacdisease. J Egypt Soc Parasitol 22, 61 1-616.

9. Badenoch, J. (1990). Cryptosporidium in water sup-plies: report of a group of experts. London: HMSO.

10. Barnard, R. J. & Jackson, G. J. (1984). Giardialamblia: the transfer of human infections by foods. InGiardia and Giardiasis, pp. 365-378. Edited by S. L.Erlandsen & E. A. Meyer. New York: Plenum Press.

sources including meats, fish, molluscs, veg-etables, or fruit, or products derived from thesefoods. Transmission is more likely to occurwhen food is eaten raw or when it has been in-completely cooked, partially pickled or smoked,or poorly preserved. Fermented foods have thepotential to be sources of infection with certainparasites, although reports of human infectionfollowing the ingestion of fermented food aresparse. Fermentation alone may therefore be in-sufficient to prevent the transmission of manyfood-borne parasites, and potentially infectedfoods should be avoided wherever possible, oralternatively subjected to either freezing or someform of heat treatment prior to processing.

11. Beaver, P. C., Gadgil, R. K. & Morera, P. (1979). Sar-cocystis in man: a review and report of five cases. Am JTrop Med Hyg 28,81 9-844.

12. Beaver, P. C., Jung, R. C. & Cupp, E. W. (1984). Clini-cal Parasitology, 9th edn. Philadelphia, PA: Lea andFebiger.

13. Beneson, M. V., Takafuji, E. T., Lemon, S. M.,Greenup, R. L. & Sulzer, A. J. (1983). Oocyst-trans-mitted toxoplasmosis associated with ingestion of con-taminated water. NEngJMed 307, 666-669.

14. Benger, A., Rennie, R. P., Roberts, J. T., Thornley J. H.& Scholten, T. (1981). A human coenurus infection inCanada. Am J Trop Med Hyg 30, 638-644.

15. Besser-Weik, J. W., Forfang, J., Hedberg, C. W.,Korlath, J. A. & Osterholm, M. T. (1996). Foodborneoutbreak of diarrhoeal illness associated withCryptosporidium parvum — Minnesota, 1995. J VetMed Assoc 276, 1214.

16. Boch, J., Laupheimer, K. E. & Erber, M. (1978). DreiSarkosporidienarten bei Schlachtrindern inSuddeutschland. Berl Muench Tieraerztl Wochenschr91,426-431.

17. Bouchier I. (1998). Cryptosporidium in water supplies:third report of the group of experts to: Department ofthe Environment, Transport and the Regions & Depart-ment of Health. Norwich, United Kingdom: HMSO.

18. Brennan, M. K., MacPherson, D. W., Palmer, J. &Keystone, J. S. (1996). Cyclosporiasis: a new cause ofdiarrhoea. Can Med Assoc J 155, 1293-1296.

19. Brodsky, R. E., Spencer, H. C. & Schultz, M. G.(1974). Giardiasis in American travellers to the SovietUnion. J Infect Dis 130, 319-323.

20. Brucker, G. (1989). Prevalence of amoebiasis and gia-rdiasis in severe intestinal disorders in interpropicalcountries. In Secnidazole: A New Approach in 5-Nitroimidazole Therapy, pp. 3—11. Edited by N. Katz& A. T. Willis. Amsterdam: Excerpta Medica.

21. Bruckner, D. A. (1992). Amoebiasis. CHn MicrobiolRev 5, 356-369.

22. Buret, A., den Hollander, N., Wallis, P. M., Befus, D.(1990). Zoonotic potential of giardiasis in ruminants. JInfect Dis 162, 23 1-237.

23. Buxton, D. (1989). Toxoplasmosis. In Pract 11, 9-12.24. Campbell, W. C. (1994). Meatborne helminth infec-

tions: Trichinellosis. In Foodborne Disease Hand-book. Vol. 2, Diseases Caused by Viruses, Parasitesand Fungii, pp. 255-277. Edited by Y. H. Hui, J. R.Gorham, K. D. Murrell & D. O. Oliver. New York:Marcell Dekker.

25. Carney, W. P., Sudomo, M. & Purnomo. (1980).Echinostomiasis: a disease that disappeared. TropGeogrMed32, 106-111.

26. Casemore, D. P. (1990). Epidemiological aspects ofhuman cryptosporidiosis. Epidemiol Infect 104, 1-28.

27. Casemore, D. P. (1991). Foodborne protozoal infec-tion. In Foodborne Illness, a Lancet Review, pp. 1 08-1 19. Edited by W. M. Waite & J. P. Arbuthnott. Lon-don: Edward Arnold.

28. Casemore, D. P., Jessop, E. G., Douce, D. & Jackson,F. B. (1986). Cryptosporidium plus Campylobacter: anoutbreak in a semi-rural population. JHyg 96, 95—105.

29. Cebotarev, R. S. & Poliscuk, V. P. (1959).Gongylonematosis of domestic animals under condi-tions of Ukrainian Polesie and forest-steppe areas. AdaParasitol Pol 7, 549-557.

30. Chen, C. Y., Hsieh, W. C., Shih, H. H. & Chen, S. N.(1987). Evaluation of enzyme-linked immunosorbentassay for immunodiagnosis of chlonorchiasis. ChungHua Min Huo Wei Sheng Wu Chi Mien I Hsueh TsaChih 20, 241-246.

3 1 . Childers, A. B., Terrell, R. N., Craig, T. M., Kaufus, T.J. & Smith, G. C. (1982). Effect of sodium chlorideconcentration, water activity, fermentation method anddrying time on the viability of Trichinella spiralis inGenoa salami. JFoodProt45, 816-819.

32. Chitwood, M. B., Velazquez, C. & Salazar, N. G.(1968). Capillaria philippensis sp. n. (Nematoda:Trichinellidae), from the intestine of man in the Philip-pines. J Parasitol 54, 368-371.

33. Cross, J. H. (1987). Public health importance ofAngiostrongylus costaricensis and its relatives.Parasitol Today 3, 367-369.

34. Cross, J. H. (1992). Intestinal capillariasis. CHnMicrobiol Rev 5, 120-129.

35. Cross, J. H. & Basaca-Sevilla, V. (1983). Experimentaltransmission of Capillaria philippinensis to birds.Trans Ploy Soc Trop Med Hyg 77, 5 1 1-514.

36. Cross, J. H. & Bhaibulaya, M. (1983). Intestinal capil-lariasis in the Phillipines and Thailand. In HumanEcology and Infectious Diseases, pp. 103-136. Editedby N. Cross & J. Cross. New York: Academic Press.

37. Cross, J. H. & Chi, J. H. C. (1982). ELISA for the de-tection of Angiostrongylus costaricensis antibodies inpatients with eosinophilic meningitis. Southeast AsianJ Trop Med Public Health 16, 110-112.

38. Daengsvang, S. (1980). A Monograph of the GenusGnathostoma and Gnathostomiasis in Thailand. To-kyo: Southeast Asian Medical Information Centre.

39. Daengsvang, S. (1982). Gnathostomiasis. In CRCHandbook Series in Zoonoses. Section C. Vol. 2, pp.800-803. Edited by M. G. Schultz. Boca Raton, FL:CRC Press.

40. Dannemann, B. R. & Remington, J. S. (1989). Toxo-plasmic encephalitis in AIDS. Hosp Med 24, 139-154.

41. Deplazes, P., Eckert, J., Pawlowski, Z. S., Machowska,L. & Gottstein, B. (1991). An enzyme-linkedimmunosorbent assay for diagnostic detection of Tae-nia saginata copro-antigens in humans. Transact R SocTrop Med Hyg 85, 391.

42. Donnelly, J. K. & Stentiford, E. I. (1997). TheCryptosporidium problem in water and food supplies.Lebensm -Weiss u-Technol 30, 1 1 1-120.

43. Doyle, P. W., Helgason, M. M., Mathias, R. G. & Proc-tor, E. M. (1990). Epidemiology and pathogenicity ofBlastocystis hominis. J Clin Micro 28, 116-121.

44. Dubey, J. P. (1993). Toxoplasma, Neospora, Sarcocys-tis and other tissue cyst-forming coccidia of humansand animals. In Parasitic Protozoa. Vol. 6, pp. 1-158.Edited by J. P. Kreir. New York: Academic Press.

45. Dubey, J. P. & Beattie, C. P. (1988). Toxoplasmosis inman (Homo Sapiens}. In Toxoplasmosis of Animalsand Man, pp. 41-60. Boca Raton, FL: CRC Press.

46. Dubey, J. P., Brake, R. J., Murrell, K. D. & Payer R.(1986). Effect of irradiation on viability of Toxoplasmagondii cysts in tissues of mice and pigs. Am J Vet Res47,518-522.

47. Dubey, J. P., Kotula, A. W., Shaar, A., Andrews, C. D.& Lindsay, D. S. (1990). Effect of high temperature oninfectivity of Toxoplasma gondii tissue cysts in pork. JParasitol 76, 201-204.

48. Dubey, J. P., Speer, C. A. & Payer, R. (1989).Sarcocytosis in humans and other primates. InSarcocystosis of Animals and Man., pp. 143-144. Ed-ited by J. P. Dubey, C. A. Speer & R. Payer. BocaRaton, FL: CRC Press.

49. Dupont, H. L. & Sullivan, P. S. (1986). Giardiasis: theclinical spectrum, diagnosis and therapy. Pediatr InfectDis 5, 131-138.

50. Duszynski, D. W. & Wattam, A. R. (1988). Coccidianparasites (Apicomplexa: Eimeridae) from insectivores.IV. four new species in Talpa europea from England. JProtozool 35, 58-62.

51. Eastburn, R. L., Fritsche, T. P. & Terhune, C. A. Jr.(1987). Human intestinal infection with Nanophyetussalmincola from salmonid fishes. Am J Trop Med Hyg36,568-591.

52. Farthing, M. J. G., Mata, L., Urrutia, J. J. & Kronmal,R. A. (1986). Natural history of Giardia infection ininfants and children in rural Guatemala and its impacton physical growth. Am J CHn Nutr 43, 395-^05.

53. Payer, R. (1981). Toxoplasmosis and public health im-plications. Can Vet J '22, 344-352.

54. Payer, R. (1994). Food-borne and waterborne zoonoticprotozoa. In Foodborne Disease Handbook. Vol. 2,Diseases Caused by Viruses, Parasites and Fungi, pp.331-362. Edited by Y. H. Hui, J. R. Gorham, K. D.Murrell & D. O. Cliver. New York: Marcel Dekker.

55. Payer, R. & Dubey, J. P. (1985). Methods for control-ling transmission of protozoan parasites from meat toman. Food Technol 39, 57-60.

56. Payer, R., Speer C. A. & Dubey J. P. (1997). The gen-eral biology of Cryptosporidium. In Cryptosporidiumand Cryptosporidiosis, pp. 1-41. Edited by R. Payer.Boca Raton, FL: CRC Press.

57. Payer, R. & Ungar, B. L. P. (1986). Cryptosporidiumspp. and Cryptosporidiosis. Microbiol Rev 50, 458-483.

58. Feachem, R., Bradley, D. J., Garelick, H. & Mara, D.D. (1983). Sanitation and Disease. Health Aspects ofExcreta and Wastewater Management. Taenia, Tae-niasis, and Cysticercosis. World Bank Studies in Wa-ter Supply and Sanitation 3. Chichester, England: JohnWiley and Sons.

59. Feldheim, W. & Knobloch, J. (1982). Serodiagnosis ofOpisthorchis viverrini infestation by an enzymeimmuno-assay. Tropenmedizin und Parasitologie 33,8-10.

60. Food and Agriculture Organization, World Health Or-ganization, Office International des Epizootics. (1992).Animal Health Year Book. Rome: Food and Agricul-ture Organization/World Health Organization.

61. Ford, P. L. & Duszynski, D. W. (1988). Coccidianparasites (apicomplexa: Eimeriidae) from insectivores.VI. six new species from the eastern mole, Scalpausaquaticus. JProtozool 35, 223-236.

62. Ford, P. L. & Duszynski, D. W. (1989). Coccidianparasites (apicomplexa: Eimeriidae) from insectivores.VII. six new species from the hairy-tailed mole,Parascalops breweri. JProtozool 35, 508-513.

63. Frenkel, J. K., Dubey, J. P. & Miller, N. L. (1970).Toxoplasma gondii in cats: faecal stages identified ascoccidian oocysts. Sd 176, 893-896.

64. Fricker, C. R. (1995). Detection of Cryptosporidiumand Giardia in water. In Protozoan Parasites and Wa-ter, pp. 91-96. Edited by W. B. Berts, D. Casemore, C.Fricker, H. Smith & J. Watkins. Cambridge, England:The Royal Society of Chemistry.

65. Galbraith, N. S., Barrett, N. J. & Sockett, P. N. (1987).The changing pattern of foodborne disease in Englandand Wales. Public Health 101, 319-328.

66. Garavelli, P. L. & Scaglione, L. (1989). Blastocystosis:an epidemiological study. Microbiol 12, 349-350.

67. Gelletlie, R., Stuart, J., Soltanpoor, N., Armstrong, R.& Nichols, G. (1997). Cryptosporidiosis associatedwith school milk. Lancet 350, 1005-1006.

68. Gilman, R. H., Brown, K. H., Visvesvara, G. S.,Mondal, G., Greenberg, B., Sack, R. B., Brandt, F. &Khan, M. U. (1985). Epidemiology and serology ofGiardia lamblia in a developing country: Bangladesh.Trans Royal Soc Trop Med Hyg 79, 469-473.

69. Guerrant, R. L. (1986). Amoebiasis: introduction, cur-rent status and research questions. Rev Infect Dis 8,218-227.

70. Habeeb, Y. S., Selim, M. A., AIi, M. S., Mahmoud, L.A., Abdel Hadi, A. M. & Shafei, A. (1996). Serologicaldiagnosis of extraintestinal sarcocystosis. J Egypt SocParasitol 26, 393^00.

71. Hadidjaja, P., Dahri, H. M., Roesin, R., Margono, S. S.,Djalins, J. & Hanafish, M. (1982). First autochthonouscase of Fasciolopsis buksi infection in Indonesia. Am JTrop Med Hyg 31(5), 1065.

72. Hall, A., Latham, M. C., Crompton, D. W. T. &Stephenson, L. S. (1981). Taenia saginata (Cestoda) inwestern Kenya: the reliability of faecal examinations indiagnosis. Parasitol 83, 91-101.

73. Harrell, L. W. & Deardorff, T. L. (1990). Humannanophyetiasis: transmission by handling naturally in-fected coho salmon (Oncorhynchus kisutch). J InfectDis 161, 146-148.

74. Haswell-Elkins, M. R., Sithithaworn, P. & Elkins, D.(1992). Opisthorchis viverrini and cholangiocarci-noma in Northeast Thailand. Parasitol Today 8, 86-89.

75. Herwaldt, B. L. & Ackers, M. L. (1997). An outbreakof cyclosporiasis associated with imported raspberries.NewEnglJMed, 336, 1548-1556.

76. Hitchcock, E. R. (1987). Cysticercosis in the UnitedKingdom. J Neurol Neurosurg Psychiatry 50, 1080-1081.

77. Hoge C. W., Shlim D. R., Rajar R., Triplett, J., Shear,M., Rabold, J. G. and Echeverria, P. (1993). Epidemi-ology of diarrhoeal illness associated with coccidian-like organisms among travellers and foreign residentsin Nepal. Lancet 341, 1 175-1 179.

78. Islam, A. (1990). Giardiasis in developing countries. InGiardiasis, pp. 235-266. Edited by E. A. Meyer.Amsterdam: Elsevier.

79. Jackson, G. J., Leclerc, J. E., Bier, J. W. & Madden, J.M. (1997). Cyclospora: Still another new foodbornepathogen. Food Technol 51, 120.

80. Jephcott, A. E., Begg, N. T. & Baker, I. A. (1986). Out-break of giardiasis associated with mains water in theUnited Kingdom. Lancet i (8453), 730-732.

81. Jiang, J. -B. & He, J. -G. (1993). Taxonomic status ofBlastocystis hominis. Parasitol Today 9, 2-3.

82. Jokipii, A. M. M., Hemita, M. & Jokipii, L. (1985).Prospective study of acquisition of Cryptosporidium,Giardia lamblia, and gastrointestinal illness. Lancet ii(8453), 487^89.

83. Kacprzak, E., Kurczewska, M. & Stefaniak, J. (1990).Two cases of cryptosporidiosis in adults. PrzeglEpidemiol 44, 245-248

84. Kain, K. C., Noble, M. A., Freeman, H. J. & Barteluk,R. L. (1987). Epidemiology and clinical features asso-ciated with Blastocystis hominis infection. MicrobiolInfect Dis 8, 235-244.

85. Karl, H., Roepstorff, A., Huss, H. H. & Bloesma, B.(1995). Survival of Anisakis in marinated herring fil-lets. IntJFoodSci Tech 29, 661-670.

86. Kirkpatrick, C. E. (1989). Giardiasis in large animals.Compend Contin EdPract Vet 1 1, 80-84.

87. Kirkpatrick, C. E. & Farrell, J. P. (1982). Giardiasis.Compend Contin EdPract Vet 4, 367-377.

88. Kliks, M. M. (1983). Anisakiasis in the western UnitedStates: four new case reports from California. Am JTrop MedHyg 32, 526-532.

89. Kotula, A. W. (1983). Postslaughter control of Tri-chinella spiralis. Food Technol 37, 91-94

90. Kum, P. N. & Nchinda, T. C. (1982). Pulmonary para-gonimiasis in Cameroon. Trans R Soc Trop Med Hyg76, 768-772.

91. Laberge, L, Griffiths, M. W. & Griffiths, M. W.(1996). Prevalence, detection and control ofCryptosporidium parvum in food. Int J Food Microbiol31, 1-26.

92. Lachberg, S., Thompson, R. C. A. & Lymbery, A. J.(1990). A contribution to the etiology of racemosecysticerocisis. J Parasitol 76, 592-594.

93. Lee, S. H., Hong, S. T., Chai, J. Y., Kim, W. H., Kim,Y. T., Song, I. S., Kim, S. W., Choi, B. I. & Cross, J. H.(1993). A case of intestinal capillariasis in Korea. Am JTrop Med Hyg 48, 542-546.

94. Levine, N. D. (1985). Chapter 6 - Amebae and Chapter13 - Ciliophora. In Veterinary Protozoology. Ames,IA: Iowa State University Press.

95. Long, E. G., White, E. H., Carmichael, W. W.,Quinlisk, P. M., Raga, R., Swisher, B. L., Dangharty,H. & Cohen, M. T. (1991). Morphologic and stainingcharacteristics of a Cyanobacterium-\ikQ organism as-sociated with diarrhoea. J Infect Dis 164, 199-202.

96. Loria-Cortes, R. & Lobo-Sanahuja, J. F. (1980). Clini-cal abdominal angiostrongylosis: a study of 116 chil-dren with intestinal eosinophilic granuloma caused byAngiostrongylus costaricensis. Am J Trop MedHyg 29,538-544.

97. Lushbaugh, W. B. & Pirtman, F. E. (1982). Amoebia-sis. In CRC Handbook Series in Zoonoses, Section C:

Parasitic Zoonoses. Vol. 1, pp. 5-13. Edited by L.Jacobs & P. Arambulo. Boca Raton, FL: CRC Press.

98. Malek, E. A. (1980). Snail-Transmitted Parasitic Dis-eases. Vol. 2. Boca Raton, FL: CRC Press.

99. Malek, E. A. (1981). Presence of Angiostrongyluscostaricensis Morera and Cespedes, 1971, in Colom-bia. Am J Trop MedHyg 30, 81-83.

100. Martin- Sanchez, A. M., Canut-Blasco, A., Rodrigues-Hernandez, J., Montes-Martinez, I. & Garcia-Rodrigues, J. A. (1992). Epidemiology and clinical sig-nificance of Blastocystis hominis in differentpopulations groups in Salamanca (Spain). Eur JEpidemiol 8, 553-559.

101. Martinez-Cruz, J. M., Bravo-Zamudio, R., Aranda-Patraca, A. & Martinez-Maranon, R. (1989). Gnathosto-miasis in Mexico. Salud Publicica Mex 3 1(4), 541—549.

102. Mason, P. & Patterson, B. (1987). Epidemiology ofGiardia lamblia infection in children: cross-sectionaland longitudinal studies in urban and rural communi-ties in Zimbabwe. Am J Trop MedHyg 37, 277-282.

103. McDonald, V. & Awad El-Kariem, F. M. (1995).Strain variation in Cryptosporidium parvum and evi-dence for distinctive human and animal strains. In Pro-tozoan Parasites and Water, pp. 104-107. Edited byW. B. Betts, D. Casemore, C. Fricker, H. Smith & J.Watkins. Cambridge, England: The Royal Society ofChemistry.

104. Meloni, B. P., Lymbery, A. J., Thompson, R. C. &Gracey, M. (1988). High prevalence of Giardialamblia in children from a WA aboriginal community.MedJAustl49,ll5.

105. Meloni, B. P., Thompson, L. C., Hopkins, R. M.,Reynoldson, J. A. & Gracey, M. (1993). The preva-lence of Giardia and other intestinal parasites in chil-dren, dogs and cats from Aboriginal communities inthe Kimberley. MedJAust 158, 157-159.

106. Meyer, V. (1965). Marinades. In Fish as Food. Vol.Ill, pp. 165-193. Edited by G Borgstrom. New York:Academic Press.

107. Millard, P. S., Gensheimer, K. F. & Addiss, D. G.(1994). An outbreak of cryptosporidiosis from fresh-pressed apple cider. JAm MedAssoc 272, 1592-1596.

108. Morera, P., Perez, F., Mora, F. & Castro, L. (1982).Visceral larva migrans-like syndrome caused byAngiostrongylus costaricensis. Am J Trop Hyg 3 1 , 67-70.

109. Murrell, K. D., Stringfellow, F., Dame, J. B., Leiby, D.A., Duffy, C. & Schad, G. A. (1987). Trichinellaspiralis in an agricultural ecosystem. J Parasitol 73,103-109.

1 10. Nansen, P. & Henriksen, Sv. Aa. (1986). The epidemi-ology of bovine cysticercosis (C. bovis) in relation tosewage and sludge application on farmland. In Epide-miological Studies of Risks Associated with the Agri-cultural Use of Sewage Sludge: Knowledge and Needs,

pp. 76-82. Edited by J. C. Block, A. H. Havelaar & P.L'Hermite. London: Elsevier.

111. Nawa, Y., Imai, J. L, Ogata, K. & Otsuka, K. (1989).The first record of a confirmed human case ofGrathostoma doloresi infection. J Parasitol 75, 166-169.

112. Nichols, G. & Thorn, B. T. (1985). Food poisoningcaused by Cryptosporidium: a load of tripe. CommunDis Rep 17, 3.

113. Nimri, L. F. (1993). Evidence of an epidemic ofBlastocystis hominis infections in preschool children inNorthern Jordan. J CHn Microbiol 3 1 , 2706-2708.

114. Nopparatana, C., Setasnban, P., Chaicumpa, W. &Tapchaisri, P. (1991). Purification of Gnathostomaspinigerum specific antigen and immunodiagnosis ofhuman gnathostomiasis. Int J Parasitol 21, 677-687.

115. Ona, F. V. & Dytoc, J. N. (1991). Clonorchis-associ-ated cholangiocarcinoma: a report of two cases withunusual manifestations. Gastroenterology 10, 831-839.

116. Ooi, W. W., Zimmerman, S. K. & Needham, C. A.(1995). Cyclospora species as a gastrointestinal patho-gen in immunocompetent hosts. J CHn Microbiol 33,1267-1269.

117. Ortega, Y. R., Sterling, C. R., Oilman, R. H., Cama, V.A. & Diaz, F. (1993). Cyclospora species: a new proto-zoan pathogen of humans. N Engl J Med 328, 1308—1312.

118. Oshima T. (1972). Anisakis and anisakiasis in Japanand adjacent area. In Progress of Medical P arasitologyin Japan. Vol. 4, pp. 301-393. Edited by K. Morishita,Y. Komiya & H. Matsubayashi. Tokyo: Meguro Para-sitological Museum.

1 19. Osterholm, M. T., Forfang, J. C., Ristinen, T. L., Dean,A. G., Washbum, J. W., Codes, J. R., Rude, R. A. &McCullough, J. G. (1981). An outbreak of foodbornegiardiasis. NEnglJMed 304, 24-28.

120. Outbreak of cyclosporiasis: Northern Virginia-Wash-ington D.C.-Baltimore, Maryland Metropolitan Area,1997. (1997). MMWR 46, 689-691.

121. Outbreaks of Esherichia coli 0157:H7 infection andcryptosporidiosis associated with drinkingunpasteurised cider — Connecticut and New York, Oc-tober 1996. (1997). MMWR 46, 4-8.

122. Pakandl, M. (1993). Occurrence of Blastocystis sp. inpigs. Folia Parasitol 38, 297-301.

123. Pape, J. W., Verdier, R. L, Boncy, M., Boncy, J. &Johnson W. D. Jr. (1994). Cyclospora infection inadults infected with HIV: clinical manifestations, treat-ment, and prophylaxis. Ann Intern Med 121, 654-657.

124. Pawlowski, Z. S. (1971). Taeniarhynchosis, aprogessive zoonosis in Europe. In: Comptes-rendus lerMultocolloque Europeen de Parasitologie, Rennes,1971, P. 35 (in French).

125. Pawlowski, Z. S. & Schultz, M. B. (1972). Taeniasisand cysticercosis (Taenia saginata). Adv Parasitol 10,269-343.

126. Pellerdy, L. (1974). Coccidia and coccidiosis, 2nd edn.Berlin, Germany: Paul Parey.

127. Petersen, L. R., Carter, M. L. & Hadler, J. L. (1988). Afood-borne outbreak of Giardia lamblia. J Infect Dis157, 846-848.

128. Phamphlett, R. & O'Donoghue, P. (1992). Antibodiesagainst Sarcocystis and Toxoplasma in humans withthe chronic fatigue syndrome. Aus NZ S Med 22, 307-308.

129. Poland, G. A., Navin, T. R. & Sarosi, G. A. (1985).Outbreak of parasitic gastroenteritis among travelersreturning from Africa. Arch Int Med 145, 2220-2221.

130. Porter, J. D. H., Gaffney, C., Heymann, D. & Parkin,W. (1990). Food-borne outbreak of Giardia lamblia.Am J Public Health 80, 1259-1260.

131. Punyagupta S., Juttijudata P., & Bunnag T. (1975).Eosinophilic meningitis in Thailand: clinical studies of484 typical cases probably caused by Angyostrongyluscantonensis. AmJ Trop Med Hyg 24, 92 1-931.

132. Qadri, S. M. H., Al-Okaili, G. A. & Al-Dayel, F.(1989). Clinical significance of Blastocystis hominis. JClin Microbiol 27, 2407-2409.

133. Ravdin, J. L, Weikel, C. S. & Guerrant, R. L. (1988).Protozoal enterpathies: cryptosporidiosis, giardiasis,and amoebiasis. In Clinical Tropical Medicine andCommunicable Diseases: Diarrhoeal Diseases, pp.503-536. Edited by R. L. Guerrant. London: BalliereTindall.

134. Rawal, B. D. (1959). Toxoplasmosis: a dye-test surveyof sera from vegetarians and meat eaters in Bombay.Trans R Soc Trop Med Hyg 53, 61-63.

135. Reimann, H. P., Mayer, M. E., Theis, J. H., Kelso, G. &Behymer, D. E. (1975). Toxoplasmosis in an infant fedunpasteurized goat milk. JPaediatr 87, 573-576.

136. Rijpstra, A. C. & Laarman, J. J. (1993). Repeated find-ings of unidentified small Isopora-like coccidia in fae-cal specimens from travellers returning to the Nether-lands. Trop Geog Med 45(6), 280-282.

137. Rommel, M. & Heydorn, A. O. (1972). Beitrage zumLebenszyklus der Sarkosporidien. llllsospora hominis(RaillietILucet, 1891)Wenyon, 1923,eineDauerformder Sarkoporidien des Rindes und des Schweins. BerlMuench Tieraerztl Wochenschr 85, 143-145.

138. Sacks, J. J., Roberto, R. R. & Brooks, N. F. (1982).Toxoplasmosis infection associated with raw goat'smilk. JAm MedAssoc 248, 1728-1732.

139. Sakanari, J. A. & KcKerrow, J. H. (1989). Anisakiasis.Clin Microbiol Rev 2, 278-284.

140. Schmutzhard, E., Boongird, P. & Vejjajiva, A. (1988).Eosinophilic meningitis and radiculomyelitis in Thai-land caused by CNS invasion of Gnathostoma

spinigerum a&& Angiostrongylus cantonensis. JNeurolNeurosurg Psychiatry 51, 80-87.

141. Schultz, M. G., Hermos, J. A. & Steele, J. H. (1970).Epidemiology of beef tapeworm infection in theUnited States . Public Health Rep 85,169-176.

142. Schulze, K. & Zimmermann, T. (1981).Sarkosporidienzysten im Hackfleisch. Fleischwirtsch61, 614-620.

143. Schwartz, D. A. (1980). Review: helminths in the in-duction of cancer: Opisthorchis viverrini, Chlonorchissinensis and cholangiocarcinoma. Tr op Geogr Med32, 95-100.

144. Shish, H. H. & Chen, S. N. (1991). Immunodiagnosisof Angiostrongyliasis with monoclonal antibodiesrecognising a circulating antigen of mol. wt. 91,000from Angiostrongylus costaricensis. Int JParasitol 2 1 ,171-177.

145. Sifuentes-Osornio, J., Porras-Cortes, G., Bendall, R.P., Morales- Villarreal, F., Reyes-Teran, G. & Ruiz-Palacios, G. M. (1995). Cyclospora cayetanensis in-fection in patients with and without AIDS: biliary dis-ease as another clinical manifestation. CUn Infect Dis21, 1092-1097.

146. Sirisinha, S., Chawengkirttikul, R., Sermswan, R.,Amornpant, S., Mongkolsuk, S. & Panyim, S. (1991).Detection of Opisthorchis viverrini by monoclonal an-tibody-based ELISA and DNA hybridization. Am JTrop Med Hyg 44, 140-145.

147. Smith, H. V. (1991). Parasitic protozoa in drinkingwater. In UK Health Related Water Microbiology ', pp.28-43. Glasgow, England: University of Strathclyde.

148. Smith, J. L. (1993). Cryptosporidium and Giardia asagents of foodborne disease. JFoodProt56, 451^61.

149. Soave, R. (1996). Cyclospora: an overview. CHn InfectDis 23, 429-^37.

150. Sterling, C. R. (1995). Emerging pathogens:Cyclospora, the Microspora, and how many more? InProtozoan Parasites and Water, pp. 243-245. Editedby W. B. Betts, D. Casemore, C. Fricker, H. Smith & J.Watkins. Cambridge, England: The Royal Society ofChemistry.

151. Sykes, T. J. & Fox, M. T. (1989). Patterns of infectionwith Giardia in dogs in London. Trans R Soc Trop MedHyg 83, 239-240.

152. Taniguchi, Y., Itashimoto, K., Ichikawa, S., Shimizu,M., Ando, K. & Kotani, Y. (1991). Human gnathosto-miasis. J Cutan Pathol 18, 1 12-1 15.

153. Taylor, M. A. (1995). Protozoan parasites and water:Veterinary aspects. In Protozoan Parasites and Water,pp. 246-249. Edited by W. B. Betts, D. Casemore, C.Fricker, H. Smith & J. Watkins. Cambridge, England:The Royal Society of Chemistry.

154. Taylor M. A. (1999). Protozoa. In The MicrobiologicalQuality and Safety of Food. Part III, Foodborne Patho-

gens, pp. 1-37. Edited by B. M. Lund, A. C. Baird-Parker & G. W. Gould. Gaithersburg, MD: Aspen Pub-lishers.

155. Taylor, M. A. & Webster, K. A. (1998). Recent ad-vances in the diagnosis of Cryptosporidium, TOKO-plasma, Giardia and other protozoa of veterinary im-portance. Res Vet Sd 65, 183-193.

156. Taylor, M. A., Catchpole, J., Marshall, R. N. & Green,J. (1993). Giardiasis in lambs at pasture. Vet Rec 133,131-133.

157. Thomson, M. A., Benson, J. W. T. & Wright, P. A.(1987). Two year study of Cryptosporidium infections.Arch Dis Child 62, 559-563.

158. Tiilsner, M. (1978). Die bestimmenden stoffichenVeranderungen im Fisch beim Salzen und Marinieren.Lebenensmittelindustrie 25 , 1 69-1 73 .

159. Ungar, B. L. P. (1990). Cryptosporidiosis in humans(homo sapiens). In Cryptosporidiosis of Man and Ani-mals, pp. 59-82. Edited by J. P. Dubey, C. A. Speer &R. Payer. Boca Raton, FL: CRC Press.

160. Upatham, E. S., Viyanant, V., Brockelman, W. Y.,Kurathong, S., Lee, P. & Kraengraeng, R. (1988). Rateof re-infection by Opisthorchis viverrini in an endemicnortheast Thai community after chemotherapy. Int JParasitol 18, 643-649.

161. Outbreaks of cyclosporiasis: United States andCanada, 1997. (1997). MMWR 46, 521-523.

162. Van Thiel, P. H. (1976). The present state ofAnisakiasis and its causative worms. Trop Geogr Med28, 75-85.

163. Verster, A. (1969). A taxonomic revision of the genusTaenia linnaeus, 1758. Onderstepoort J Vet Res 36, 3-58.

164. Walsh, J. A. (1986). Problems in recognition and diag-nosis of amoebiasis: estimation of the global magnitudeof morbidity and mortality. Rev Infect Dis 8, 228-238.

165. Walther, M. (1980). Taenia saginata cysticercosis: acomparison of routine meat inspection and carcass dis-section results in calves. Vet Rec 106, 401^02.

166. Walzer, P. D. & Healy G. R. (1982). Balantidasis. InCRC Handbook Series in Zoonoses, Section C: Para-sitic Zoonoses. Vol. 1, pp. 15-24. Edited by L. Jacobs& P. Arambulo. Boca Raton, FL: CRC Press.

167. Walzer, P. D., Judson, F. N., Murphy, K. B., Healy, G.R., English, D. K. & Schultz, M. G. (1972). Balantidi-asis outbreak in Truk. Am J Trop Med Hyg 22, 33^4 1 .

168. Warhurst, D. C. & Green, E. L. (1988). Protozoalcauses of diarrhoea. PHLS Microbiol Dig 5, 31-37.

169. Welch, J. S., Dobson, C. & Campbell, G. (1976).Immunodiagnosis and seroepidemiology ofAngiostrongylus costaricensis zoonoses in man. TransR Soc Trop Med Hyg 74, 614-623.

170. White, K. E., Haedberg, C. W., Edmondson, L. M.,Jones, D. B. W., Osterholm, M. T. & McDonald, K. L.

(1989). An outbreak of giardiasis in a nursing homewith evidence for multiple modes of transmission. J In-fect DiS9 160,298-304.

171. Work, K. (1968). Resistance of Toxoplasma gondii en-cysted in pork. Acta Pathol Microbiol Scand 73, 85-92.

172. World Health Organization. (1979). Report of a WHOExpert Committee, with the Participation of the FAO.(Technical Report Series 637).Geneva: Author.

173. World Health Organization. (1983). Guidelines forSurveillance, Prevention and Control of Taeniasis/Cysticerocosis . Geneva: Author.

174. Wurtz, R. (1994). Cyclospora: a newly identified intes-tinal pathogen of humans. Clin Infect Dis 18, 620-623.

175. Wyllie, A. S. (1984). Cryptosporidiosis. Br Med J 2^9,1383-1384.

176. Yii, C-Y. (1976). Clinical observations on eosinophilicmengingitis and meningoencephalitis caused byAngiostrongylus costaricensis on Taiwan. Am J TropMed Hy >g 25, 233-249.

177. Yilmaz, S. M. & Hopkins, S. H. (1972). Effects of dif-ferent conditions on duration of infectivity of Toxo-plasma gondii oocysts, J Parasitol 58, 938-939.

178. Zierdt, C. H. (1991). Blastocystis hominis: past and fu-ture. Clin Microbiol Rev 4, 61-79.

179. Zinter, D. E. & Migaki, G. (1970). Gongylonemapulchrum in tongues of slaughtered pigs. JAm Vet MedAssoc 157(3), 301.

CHAPTER 10

Biotechnology and Food Safety:Benefits of Genetic Modifications

T. Verrips

INTRODUCTION

A number of consumer research studies haveshown that fermented food products are re-garded by consumers as healthy and natural.9

Factually and from a historic point of view, thisis correct. During the approximately 7,000 yearsthat fermented foods have been produced, anenormous selection process has taken place, andthe tasty products that are presently on the mar-ket are the result of this selection process. It iswell known that fermentation of foods anddrinks prolongs the shelf life of these products.The historic claims that fermented foods, par-ticularly those derived from milk, prevent dis-eases are supported by many studies. Thoseclaims related to mucosal health are excellentlysummarized by Salminen et 0/.23 Nevertheless,the title of this chapter suggests that at present,there is concern about the safety of fermentedfood products. This is due to the rapidly growingapplication of genetically modified food prod-ucts, including fermented food products. In par-ticular, the penetration on the market of geneti-cally modified plants is remarkable (Figure 10-1and Table 10-1). In addition, more than 20 foodenzymes that are on the market at present areproduced using recombinant DNA technology(Table 10-2).

The perception of genetic modification offood products by European consumers is notvery positive and varies from country to country.Generally, the attitude of Western consumers isneutral to positive (Figure 10- 2), whereas in

most of the developing countries, this technol-ogy is seen as an opportunity.20 However, inspite of the present problems in a number ofWest European countries, and excluding smallgroups of dogmatic opponents of genetic engi-neering, the perception of the consumer will ulti-mately be determined by the benefits versusrisks ratio.

In spite of many efforts, neither scientists noropinion makers have been able to properly ex-plain the potential risks of genetically modifiedfood products. Therefore, many consumers per-ceive recombinant DNA technology as an intrin-sically dangerous technology, although in the 25years that this technology has been applied, nounintended dangerous materials have been pro-duced. The main reason for the present percep-tion of this technology by West European con-sumers is that in many discussions, no cleardefinitions of the nature of the genetically modi-fied food are used. In an attempt to rationalizethe discussions and to avoid inappropriate ethi-cal discussions, let us examine simple decisionschemes based on rules of the U.S. Food andDrug Administration (FDA), the United King-dom, and the Netherlands,29 and a model forvarious genetically modified products (Figure10-3).30 It should be emphasized that this figurereflects the author's personal view concerningthe safety of genetically modified foods that donot contain (antibiotic) resistance markers("clean"). Genetically modified plants or micro-organisms containing antibiotic resistance mark-ers present a very difficult to quantify but realis-

tic risk of spreading genes encoding antibioticresistance in the environment.6'10 The chancethat these distributed antibiotic-resistant genesare taken up is very low, but not zero, and there-

Table 10-1 Rapid Increase of TransgenicPlants

Crop 1996 1997

Maize 525,000 4,400,000Soybean 400,000 5,250,000*Potato/tomato 40,000 500,000+Oilseed 200,000 1,600,000Cotton 810,000 1,200,000

"Includes the South Amerian 1996/1997 harvest.+Includes estimates for China.Note: In 1994, no transgenic crops were cultivated for

usage in the agrofoods industry.

fore undermines the argument that geneticallymodified plants or microorganisms containing agene that encodes an intrinsically safe proteinare safe.

In addition to the lack of proper explanationof the potential risk of genetically modifiedfoods, the benefits for the consumer are eithernonexistent or not communicated properly toconsumers. This is a major weakness, and onlywhen the benefits to consumers become clearwill genetically modified foods be accepted bythe majority of consumers. It is obvious that ben-efits are not absolute values but relative values,strongly influenced by demographic, geo-graphic, and socioeconomic factors.

In this chapter, the supply chains of fermentedfoods are taken as guides to point out the poten-tial benefits and risks of every step of thesechains. The emphasis will be on fermented foods

first GMO first GMObacterium plant

Note: X-axis-penetration as a function of time: The expression of somatostatin in E. coll is taken asthe starting point and insulin as the first large-scale pharmaceutical product. Y-axis-estimatedpercent of rDNA product in certain area (Detergent enzymes are nearly 100% rDNA products, about25% of the food enzymes (e.g., chymosin) are made via rDNA technology).

Figure 10-1 Penetration of rDNA technology in agriculture, Pharmaceuticals, and consumer products.

SomatostatinInsulin

Foodenzymes

7 Qualitytrait crops

GMOMicroorganisms

Agronomictrait crops

Agronomictrait cropsDetergent

enzymes

Estimatedsales GMO/sales total

Table 10-2 Commercial Food Enzymes Made by rDNA Technology (as on the Market at 1.5.1998)

Enzyme Commercial Name Producer Application

oc-Galactosidase cc-galactosidase NOVO N. Animal feedXylanase Bio-feed wheat id idLipase Lipozym id lnteresterificationALDC Maturex id BeerAmylase Novamyl id Bread1,3 Lipase Novozym 677 id BreadXylanase Peptopan id BreadAmylase Maltogene id StarchChymosin Novoren id CheeseLipase Novozym 398 id lnteresterificationLipase Novozym 435 id idFytase Pytan id Animal feedAmylase Termamyl id StarchTransferase Toryzym id StarchChymosine Maxiran DSM/Gb CheesePhytase Natuphos id Animal FeedChymosin Chymogen C. Hansen CheeseChymosin Chymax Pfizer CheeseXylanase Biobake710 Quest BreadInvertase id ChocolateGlucanase id Beeroc-Galactosidase id Guar modification

starting from traditionally bred or geneticallymodified plants and milk from traditionally bredcattle. The approach of analyzing the safety of fer-mented products as a function of the supply chainhas been adopted, because only with such an ap-proach will all potential risks be included, which isessential to gain the confidence of consumers.

By analogy with microbial risk assessments,28

risk is defined as the probability of a hazard oc-curring and hazard as a harmful event. Thismeans that when the newly introduced gene en-codes a protein that is intrinsically safe and doesnot provide a selection benefit for any receivingorganism, the hazard is considered as zero. Con-sequently, the risk is zero and the safety of thefood product is 100%. If, on the other hand, thenewly introduced gene encodes for a protein thatis intrinsically safe but provides the receivingorganisms with a selection benefit in the openenvironment, then the health hazard is zero, butthe environmental hazard is not zero. Conse-

quently, the probability of transfer of the gene tothe receiver has to be estimated. Genes encodingintrinsically unsafe proteins (i.e., health hazard> O) are not considered because authorities willnot give permission for cloning of such genes inplants and/or microorganisms that enter the foodchain; neither will the food industry ever usesuch organisms.

SUPPLY CHAIN OF FERMENTEDFOODS AND ITS IMPORTANCE TOENSURE SAFE PRODUCTS

Table 10-3 provides a selection of fermentedfoods based on the raw material(s) that is (ormay become in the near future) derived from ge-netically modified plants.

The steps in the supply chains (including po-tential beneficial and risk aspects) of fermentedfoods using traditionally bred plant materials ormilk from traditionally bred cows with fermen-

Figure 10-2 Acceptance of rDNA technology for consumer products derived from plants and microorganismsin various countries. Acceptance varies from approximately 90% in Canada, United States, and Portugal to 30and 45% in Austria and Germany, respectively.

tation processes using genetically modified mi-croorganisms are provided in Figure ICM. How-ever, Tables 10-1 and 10-3 illustrate that geneti-cally modified plants will become a majorsource of plant raw material for fermented foodsin the near future, and therefore a separate sup-ply chain risk/benefits analysis for fermentedproducts derived from genetically modifiedplants is provided in Figure 10-5.

It is impossible to deal in this chapter with all ofthe products in Table 10-3 for both traditional andgenetically modified raw materials and/or for tra-ditional and genetically modified microorganismsused in the fermentations. Therefore, only two ex-amples are worked out in more detail.

1 . soy sauce — a fermented product using ge-netically modified wheat and/or soybeans,traditional or genetically modified micro-

organisms, and traditional or recombinantDNA enzymes

2. cheese — a fermented product using nor-mal cow's milk, genetically modifiedfunctional microorganisms, and/or en-zymes made by genetically modified mi-croorganisms

These two examples cover quite well thewhole spectrum of fermented food products inwhich recombinant technology is used at somestep in the overall process.

As stated in the introduction to this chapter, fer-mented foods derived from genetically modifiedplant material or produced by genetically modi-fied microorganisms will only be accepted if thereare clear direct (e.g., significantly better quality orshelf life, more healthy or more convenient) or in-direct (e.g., availability, environment) benefits for

SpainItaly

PortugalFrance

BelgiumIreland

UK

CanadaNorway

NetherlandSwedenFinlandJapan

DenmarkAustria

GermanyUSA

Known Unknown Knownnonintegrated integrated integrated

Note: X-axis-physical state of foreign gene(s): "known" means that the location of foreign gene(s) is/areknown exactly, "integrated" means stable integration of foreign gene(s) on the chromosome of the host;Y-axis-host organism; Z-axis-whether the consumer product is free of DNA (e.g., cheese made withchymosin) or product contains inactivated cells that produced the rDNA product (i.e., tomato paste) orthe rDNA producing cells are "alive" present in the product (i.e., fresh tomatoes). Vertical spheres-norisk; white spheres-risk assessment inconclusive; dotted spheres-either risk assessment inconclusiveor ethical objections against such consumer products.

/Vote: This scheme is based on the assumption that the used constructs do not contain an antibioticresistance marker. Otherwise, the risk to the environment will increase for all cases except containedfermentation of microorganisms.29

Figure 10-3 Summary of risk assessment of rDNA products on basis of three parameters.

the consumer. Some potential benefits will be dis-cussed in the following sections.

Some Consumer Benefits of GeneticallyModified Plants

Benefits are relative, and for each group ofconsumers, benefits will be perceived differ-ently. Consumer studies in Western Europeshow that the order of consumer criteria to buyproducts is quality, health aspects, convenience,environment, and price. It is likely that price to-gether with availability will be on top of such alist in developing countries. In the Westernworld, there is a surplus of food raw materials

and improvement of crop yield is not perceived asa benefit by the consumer; neither is the loss ofmaterial during transport seen as a problem. How-ever, the maintenance of quality during transportand at home is considered as a benefit by consum-ers, as was demonstrated by the initial successfulintroduction of the Flav Sav tomato (with anantisense polygalacturonase gene that ensured de-lay of softening of the tomato) by Calgene on theUnited States market in the mid-1990s.

In many developing countries, however, thereis at present a shortage of food raw materials.Therefore, an increase in the yield of crops bygenetic modification is seen as a major advan-tage. There is evidence that shows that in the

Host

Microorganism

Plant

Animal

Type of product

DNA in intact cell

DNA in dead cell

Free of DNA

Table 10-3 Overview of Most Important Fermented Food Products Derived from Plants of which Ge-netically Modified Varieties Are on the Market and the Main Functional Microorganisms

Raw Materials

Cassava

/ sorghum

Maize

/ cassavaRice

/ Black gram

/ Carrots/ Soybeans

or cassavaSoybean

/ rice

/ wheat

/ wheat

/ wheatWheat

/milk

Functional Microorganisms

BacteriaCorynebacterium,GeotrichumLactic acid bacteria,Saccharomyces, CandidaAspergillus, lactic acidbacteria, yeastYeasts, bacteriaLactic acid bacteria,yeasts, moldsLactic acid bacteria, yeastsMonascusRhizopusLactic acid bacteriaSaccharomycesAspergillus, BacillusHansenula, Candida,GeotrichumLactic acid bacteriaTorulopsisHansenulaAspergillus, Lactic acidbacteria, yeastsYeasts, moldsMucor, AspergillusAspergillus, RhizopusActinomucorBacillus nattoActinomucor, MucorRhizopusAspergillus, yeastsLactic acid bacteriaAspergillus, Lactic acidbacteriaAspergillus, yeastsLactic acid bacteriaAspergillusSaccharomycesMoldsLactic acid bacteria

Product

Chickwangne (Congo)

Gari (Nigeria)

Burukutu (Nigeria)

Chicha (Peru)Jamin-bang (Brazil)

Ogi (Nigeria)Banku (Ghana)Ang-kak (China)Lao-chao (China)

Puto (Philippines)Sierra rice (Ecuador)

Torani (India)

IdIi (India)Kanji (India)

Miso (China, Japan)Tape (Indonesia)Chee-fan (China)Meju (Korea)Meitauza (China)Natto (Japan)Sufu (China)Tempeh (Indonesia)

Miso (China, Japan)

Hamanatto (Japan)

Soy sauce (southeast Asia)Tao-si (Philippines)Jalebies (India, Pakistan)Minhin (China)Kishk (Egypt)

next 10 years, the increase in the supply of agri-cultural products will be less than the expectedincrease in world population and food shortage,especially in developing countries, will in-crease16 (Figure 10-6). It has also been sug-

gested that only when modern biotechnology re-sults in a second green revolution may this glo-bal problem be prevented.17

Also, the prevention of the substantial lossesduring transportation of plant raw materials (up

Figure 10-4 Schematic benefit versus risk ratio of rDNA products made by microorganism.

to 30% of harvested plant materials is destroyeddue to microorganisms, insects, or uncontrolledendogenous processes) is seen as a major benefitfor consumers in that part of the world.

There is evidence that modern biotechnologycan significantly contribute to increasing yieldsor reducing losses of plants and plant products inthe first phases of the supply chain. Monsanto 'sRoundup herbicide-resistant crops need less her-bicides (approximately 10% less) and show in-creased yields (5-15%) (Farmers OrganizationArgentina, personal communication, November1997).

Plants producing Bacillus thuringiensis BT-protein (preferably more than one variety of this

protein so as to minimize the probability of ad-aptation) are quite well protected against insects,thereby increasing the yield. Modern biotech-nology can also contribute to the reduction oflosses due to microbial spoilage during distribu-tion. Plants have a rich arsenal to defend them-selves against invading microorganisms.5 Theycan change the structure of plant tissue by exten-sive cross linking catalyzed by redox enzymes,produce low molecular mass antimicrobialagents, and degrade cell walls of invading mi-croorganisms using hydrolases they produce. Inparticular, plant pathogenic fungi can be effec-tively destroyed by these enzymes. Some com-panies like Novartis and Zeneca are very active

Environment

Consumption

Distribution

Raw materials

Processing, includingfermentation

Faster process,less salt

Price reduction,better quality, healthier

See 2.2 a-d and3.1, step 1a, 1b

ZEROSee 3.1, step

2a-c

Benefits Risks

Figure 10-5 Schematic benefit versus risk ratio of rDNA product made by plants.

in this field, as shown by the five and seven pat-ents they filed respectively on this technologybetween 1980 and 1997. This approach is lim-ited to improving existing plant protection sys-tems. However, recombinant DNA technologycan do more. Plant Genetic Systems had aproject in the mid-1980s to express potent anti-microbial peptides (e.g., apidaecin)3 in plants,although with the intention to isolate these anti-microbials from the plant and use them in alltype of products. At present, a large number ofantimicrobial peptides are known, some ofamazing effectivity.1

A consumer benefit of currently available ge-netically modified plants can be a lower price dueto higher yield, reduced cost for chemicals, andreduced losses during harvesting and transport.

However, an important benefit can be delivered byplants with the right balance of micronutrients thatcontribute to consumer health. Plants are a well-known and extremely rich source of all kinds ofcomponents that may contribute to consumerhealth, such as antioxidants, organic metal com-plexes, phytoestrogens, phytosterols, and vita-mins. At present, the diet of approximately 2.5 bil-lion people contains insufficient amounts ofminerals and vitamins. The consequences of theseshortages are dramatic. For example, each year, 2million young children die and approximately300,000 children go blind due to a shortage of vi-tamin A. With the rapid increase in our knowledgeof cell and molecular biology, and much better andfaster analytical techniques, the number and kindof plant and microbial components with sustain-

Environment

Consumption

Distribution

Processing, includingfermentation

Raw materials

See 3.2,step 2a-2b

See 3.2, step 3

See 3.2, step 4

See 3.2, step 5

See Table 10-2

Benefits Risks

Year

Figure 10-6 World grain harvest during the last 35 years. A steady increase in total production from 1.2 to 2.2billion of metric tons; a rapid increase in amount per capita from 1966 to 1987, followed by a decrease from 1988to present.16 Source: Reprinted with permission from FAOSTAT, C.C. Mann, Crop Scientists Seek a New Revo-lution, Science, Vol. 283, pp. 310-314. Copyright 1999 American Association for the Advancement of Science.

able health claims will increase significantly in thenear future.

What happens to these components with po-tential healthy properties during fermentation isnot very well known, but it is likely that chemi-cal nature and bioavailability of these compo-nents will be influenced (for better or worse) bythe various steps in a fermentation process.Some literature is available on the conversion ofestrogens and phytoestrogens by the microfloraof the gastrointestinal system and show thatphytoestrogens can be (extensively) convertedinto more effective compounds, such as estroneinto oestradiol.21 However, knowledge in thisarea is too fragmented and limited, and this im-portant aspect has to be researched.

Some Potential Hazards and Risks ofGenetically Modified Plants

It is outside the scope of this chapter to go intodetail on the potential hazards of genetically

modified plants. However, the integral supplychain approach followed requires that some as-pects of the potential risk are discussed. Atpresent, there are at least four issues in relationto genetically modified plants.

1. Most of the genetically modified plantscontain an antibiotic resistance gene asmarker and, although the marker gene assuch is not expressed, the potential transferof this marker gene to other crops in thesurrounding area has been studied.11 Al-though the frequency of such an event islow, it is measurable; therefore, this ap-proach may contribute to a further spread ofantibiotic resistance genes in the environ-ment, thereby contributing to an extremelyundesirable further increase of antibioticresistance of human and animal pathogens.

2. The gene providing the plant with a new de-sired property, such as herbicide resistance,can also be spread in the environment. This

Kilo

gram

s/ca

pita

World grain harvest per capitaP

roduction (billions of metric tons)

means that there is a small probability thatweeds will pick up this property and be-come resistant to herbicides. This spread ofgenes providing herbicide resistance is anissue that has to be addressed.

The spread of genes providing plantswith new desired properties related tohealth benefits of consumers will also oc-cur, but in this case, the gene will mostprobably not provide the recipient with anecological benefit; therefore, this risk is solow that it can be neglected.

3. It is still not possible to integrate the newgenetic information in a predeterminedplace in the genome of the plant. The ran-dom integration may result in the destruc-tion or enhancement of the expression ofgenes at the locus of integration and there-fore may change the metabolism of theplant. To exclude the introduction of a haz-ard, one should precisely determine the po-sition of integration and use techniques suchas DNA microarrays,4 proteomics,13 and gaschromatography coupled to mass spectrom-etry (GC/MS) analysis of metabolites to de-termine the effects of this random integra-tion on the metabolism of the plant.

4. It is possible that pieces of plant DNA aretaken up by the epithelial cells of the gas-trointestinal tract (GIT) of consumers.Hard data on this transfer are scarce, butthis is a general phenomenon of digestionand uptake of foods and not just an issuerelated to genetically modified plant mate-rial. However, particularly when antibi-otic-resistant markers are used in geneti-cally modified organisms (GMOs), it isnecessary to determine the probability ofthis transfer in order to make a good riskassessment. Of course, it would be muchbetter if antibiotic resistance markers werenot present at all.

Some Consumer Benefits of GeneticallyModified Microorganisms

As stated in the introduction to this chapter,the popularity of fermented food is based on sev-

eral aspects, of which the enormous variation inappearance, flavor, and texture is the most im-portant for Western consumers. This enables theconsumer to select a product out of this enor-mous range that is close to personal preferenceand not an "average" product. When fermentedfoods are made from plant materials that do nothave an optimum composition from a nutritionalpoint of view, such as cassava, fermentationcontributes to (partial) supplementation of thesecomponents (e.g., amino acids such as methion-ine, lysine, and tryptophan; vitamins, in particu-lar A and B vitamins; minerals). Fermentationand the accompanying physical processes alsocontribute to the elimination of undesirablecomponents that are present in plants2 (seeChapter 4). In general, fermented products alsohave a better microbiological stability, which isa considerable consumer benefit, especially indeveloping countries with less well-organized(chilled) distribution chains and often no refrig-erator in the homes (see Chapter 2).

On top of the general benefits for the consumerdescribed above, certain microorganisms used infermentation processes, particularly lactic acidbacteria (LAB), may provide additional healthbenefits (Exhibit 10-1). Recombinant DNA tech-nology may be able to improve these positive as-pects of fermented food products. It is even pos-sible to extend this range of potential healthbenefits to consumers, for example, by oral immu-nization against pathogens and/or toxins.31

Some Potential Hazards and Risks ofGenetically Modified Microorganisms orTheir Enzymes Used in theManufacturing of Fermented Foods

There are several aspects related to the safetyof fermented foods. Assuming that extensive re-search and risk assessment have proved that thenewly introduced gene produces an intrinsicallysafe protein and that this protein does not con-tribute to the biogenesis of any harmful compo-nent, either during fermentation or during diges-tion of the fermented product, the remainingsafety aspects are:

Exhibit 10-1 Potential Consumer Benefits ofFermented Foods

• Improved appearance, flavor, and texture• Conversion of antinutritional and toxic com-

pounds (e.g., removal of cyanogenic glyco-sides in cassave)

• Increased microbiological keepability• Enrichment with amino acids, minerals, and

vitamins• Agonistic action against enteric pathogens*• Improved lactose utilization (important for

developing countries)*• Conversion of potential precarcinogens into

less harmful compounds*• Stimulation of the mucosal immune systemf

*Mainly related to foods fermented with lacticacid bacteria.

fOnly a small number of lactic acid bacteria mayhave this property.

• Can the transfer of newly introducedgene(s) to recipient microorganisms occurin the fermentation process or after con-sumption of the product in the GIT of con-sumers?

• Can the transfer of newly introducedgene(s) to recipient cells of the GIT of con-sumers occur?

• Can the transfer of newly introducedgene(s) to recipient microorganisms in theenvironment occur?

The risk assessment of the aspects mentionedunder the first and third bullet have been de-scribed in some detail earlier.30 As the potentialfor transfer of genetic information from donormicroorganisms to recipient microorganisms isa matter of concern, and new data are available,this will be discussed in some detail using thedecision tree provided in Figure 10-7.

Provided that the microorganism used in a fer-mentation process is intrinsically safe and therecombinant version of this microorganism issubstantially equivalent and free of any DNA ofthe marker free (clean) host organisms, the risk

related to the use of these recombinant DNA mi-croorganisms during fermentation processes iszero for both consumer and environment.30

When, during the fermentation process, achemical or physical treatment is applied that Iy-ses the genetically modified microorganism anddestroys the functionalities of the genome, it isnot necessary to take into account the aspect ofpotential transfer of genetic material from thesemicroorganisms to recipient cells in the GIT ofconsumers or safety aspect (third bullet above).

TWO SETS OF EXAMPLES OF SAFETYEVALUATION OF FERMENTEDFOODS FOR WHICH RECOMBINANTDNA IS USED

Supply Chain for Soy Sauce Produced fromGenetically Modified Soybeans and/orWheat and/or Using Genetically ModifiedMicroorganisms and/or EnzymesProduced by Genetically ModifiedMicroorganisms

There are a large number of fermented prod-ucts that are derived from wheat and/or soybean(see Chapter 1). For the supply chain risk/ben-efits analysis, the well-known soy sauce wasused as an example. A general outline of thebenefits and risks of genetically modified soy-beans was provided on pages 223 and 227, andfor microorganisms and enzymes used in the fer-mentation process, on pages 228. This outlinewill now be worked out in detail for soy sauce. Asimplified scheme of the soy sauce process isdepicted in Figure 10-8.

The safety of soy sauce that is derived from ge-netically modified soybean will be analyzed as afunction of the supply chain. As pointed out onpage 227, a genetically modified soybean shouldnot have an antibiotic resistance marker becausethis will increase the risk and is of no benefit to theconsumer. For this example, it is assumed that thegenetic modification of the soybean results in afaster fermentation process due to the incorpora-tion of enzymes in the soybean and/or wheat thatcontribute in the first steps of the process. Thefaster process is assumed to also contribute to thesensory quality of the soy sauce. Therefore, the

benefits for the consumer will be better quality anda price reduction.

• Step Ia (Figure 1(M). The risk related tothe first step of the supply chain (cultivationof the soybeans): Because the newly intro-duced gene encodes an intrinsically safe pro-tein that does not provide any receiving or-ganism with an ecological advantage, boththe environmental and consumer health risksare zero. It should be noted, however, that theenvironmental hazard is not zero. Afterstudying the frequencies of spontaneous hy-bridization between oilseed rape (Brassicanapus) and weedy rape (B. campestris)Jorgensen & Anderson11 concluded thattransgenes from oilseed rape may be pre-

served for many years, and that weedy B.campestris with transgenes may present eco-nomic (ecological) risks to farmers andplant-breeding companies or biochemical in-dustries. Although their findings of sponta-neous hybridization have been confirmed inother studies, unfortunately, they mixed uprisk and hazard. Only when the transgene re-sults in a clear ecological advantage for therecipient is there a hazard, and, based on theirstudies, a realistic probability (> 1 0~8) that thehazard occurs (= risk). Moreover, the transferof genetic material has happened for millionsof years and has not resulted in ecologicalproblems because the transgene did not givethe recipient an advantage under a large num-ber of ecological conditions.

1(E). Generally recognized as safe ge-netically modified microorganismused in fermentation and still livingin the final product

2(Q). Encode the newly introduced genefor an intrinsically safe product?

3(Q). Is the new gene stable integrated inthe genome of the host cell?

4(Q). Is an antibiotic resistance markerused during construction of theGMO?

5(Q). Does the newly introduced gene en-code for a protein that may resultunder certain conditions in an eco-logical advantage for the host cell?

6(A). Carry out full risk assessment.7(Q). Does the newly introduced gene en-

code for a protein that after transfermay result under certain conditionsin an ecological advantage to recipi-ent mammalian cells?

8(Q). Does the newly introduced gene en-code for a protein that after transfermay result under certain conditionsin an ecological advantage to recipi-ent microbial cells?

9(A). Integrate new gene, preferably onpredetermined locus on the chromo-some of the host.

10(Q). Is the antibiotic resistant markergene eliminated?

(E). Entry or end; Diamonds(Q = questions); Y = yes, N = No

Figure 10-7 General scheme to determine safety to consumers and some environmental consequences of geneti-cally modified microorganisms used in fermentation processes. Note: This is a modification of the schemespublished in Verrips and Vandenberg.30

STOP

STOP

STOP

No safety orecological hazard

It is also important to consider the stabil-ity of plant DNA in soil because that DNAcan be picked up by microorganisms thatare present in the soil. A study showed that"survival" of a neomycin/kanamycin resis-tance gene from transgenic tobacco was0.14% after 120 days in soil. Most prob-ably, the persistence is so high because ofstabilization of the DNA by soil.26 Becausethe stability of DNA in soil is higher thanwas assumed in the past, the probabilitythat DNA is taken up by soil organisms isalso higher. Indeed, various studies haveshown that horizontal transfer of plantDNA to microorganisms can occur, but atlow frequencies. For instance, based on anumber of well-controlled experiments, thetransfer of DNA from transgenic potato tothe plant pathogenic bacterium Erwiniachrysanthemi under natural conditions iscalculated to be 10~17, which is much lowerthan the detection limit of approximately10-12 24 Eariier experiments showed that thetransfer of hygromycin resistance fromvarious transgenic plants (B. napus, B. ni-

gra, Datura innoxia, and Vicia narbonen-sis) to the fungus Aspergillus niger aftercocultivation occurred with an unexpect-edly high frequency.10

• Step Ib (Figure 10-4). In cases where thenewly introduced gene encodes a proteinthat can provide an ecological advantagefor any recipient organism, the risk is notzero and the risk assessment (Figure 10-7)should be performed before proceeding tostep 2.

• Step 2a (Figure 10-4). The risk related tothe second step of the supply chain (fer-mentation process): During this process,the soybean is physically denatured; more-over, the chemical composition in certainparts of the process is very hostile for or-ganisms, with the exception of some func-tional organisms.

• Step 2b (Figure 10-4). In the cases that thesoy sauce process contains geneticallymodified microorganisms, it is very likelythat the DNA of these microorganisms willalso be destroyed during the final steps ofthe process when the conditions are quite

Figure 10-8 Schematic outline of soy sauce process.

bottling soy sauce

pasteurization

infusion/flotation

fermentation (30-40 0C, NaC113%, 10 days)

yeast/LAB seedcultures

hydrolyses (40-50 0C, NaCI 8%, 10 days)

mold seed culturekoii

autoclaving

sugar/salinesolution

liquefactionsaccharification

syrup

wheat flour/rice brinedef. soy bean

soaking in water

wheat bran

brine

harsh (but hard evidence is missing). It isagain highly recommended to avoid thepresence of any nonfunctional foreigngenes or genes encoding antibiotic resis-tance properties in these microorganisms.The hazard of the newly introduced geneproduct should be analyzed in the sameway as described in step Ia for new genesintroduced in soybeans and/or wheat.

• Step 2c (Figure 10-4). In the case that en-zymes produced by genetically modifiedmicroorganisms are used in the soy sauceprocess, the regulations require that the en-zyme should be intrinsically safe and freeof DNA of the microorganism involved.29

Practice proves that this requirement can befulfilled without much difficulty in indus-try. Even though there is evidence that en-zymes can be made free of any DNA, as ad-vocated earlier, it is highly recommendedto produce enzymes only with food grademicroorganisms that are free from non-functional foreign genes, in particular, freefrom genes encoding antibiotic resistanceproperties. The enzyme as such should be"substantially equivalent" to the wild typeenzyme, another requirement that can bemet by industry. Consequently, using re-combinant DNA enzymes in the soy sauceprocess does not pose any hazard and there-fore no risk to consumer and environment.

• Steps 3-5 (Figure 10-5). In all of the sub-sequent steps of the supply chain, includingthe step in which consumers introduced the(digested) soy sauce in the environment,the risk related to the use of geneticallymodified soybean and/or wheat or the useof genetically modified microorganisms orenzymes produced by recombinant DNAtechnology remains zero.

Supply Chain for Cheeses Produced withGenetically Modified LAB and/orEnzymes Produced by GeneticallyModified Microorganisms

As stated in the introduction, fermented prod-ucts, particularly fermented milk products, have

a healthy image. Although a number of studiesstill have to be carried out to substantiate the po-tential health benefits of LAB, in particular, theeffects of LAB and their products on mucosalhealth, there are already a large number of publi-cations that support these benefits for consum-ers.23 The most important potential health andother benefits are summarized in Exhibit 10-1 .

Some of the main aspects of cheese processesare provided in Figure 10-9, and the risk assess-ment of that process and the other steps in thesupply chain will be performed similarly to theassessment for soy sauce.

• Step 1 (Fig 10-4). Traditional milk will beused for cheese production; therefore, a riskassessment of GMO-related issues in thisstep is not necessary.

• Step 2a (Fig 10-4). In this step, geneticallymodified LAB are introduced in the cheeseprocess. As the safety of the end product(cheese) for consumers will be discussed instep 4, only the environmental safety aspectsare discussed in this step. It is unrealistic toconsider the cheese fermentation process as acompletely closed process; therefore, there isa probability that genetically modified LABwill enter the environment. Assuming theseLAB do not contain any foreign, nonfunc-tional genes, and that the newly introducedgene encodes an intrinsically safe protein, thehazard will be zero and consequently the riskwill also be zero.

If, however, the LAB contain a gene-en-coding antibiotic resistance or a propertyproviding an ecological advantage to therecipient, the environmental hazard is notzero and a formal risk assessment as dis-cussed in detail in an earlier publication30

should be performed. More recent data15'18

support the view that horizontal transfer ofgenetic material from one microorganismto another, either directly by conjugation ortransformation or indirectly via transduc-tion, occurs at a much higher frequencythan was assumed in the mid-1980s.

• Step 2b (Figure 10-4). In this step, en-zymes made by recombinant DNA technol-

ogy are introduced.27 As described in thesoy sauce example, an intrinsically safe en-zyme produced under proper conditionsdoes not pose a risk to consumer or envi-ronment.

• Step 3 (Figure 10-5). During distribution,the probability of the escape of LAB intothe environment is extremely low, and ashorizontal transfer also occurs with low fre-quencies, this probability is considered tobe effectively zero.

• Step 4 (Figure 10-5). In this step, the con-sumer will be in direct contact with the ge-netically modified LAB and/or the enzyme.For the enzyme, a relatively straightfor-ward procedure has to be followed to provethat the enzyme is intrinsically safe, origi-nates from a well-known source, and issubstantially equivalent to the wild type en-zyme.29 If that is the case, the enzyme willnot pose a hazard and consequently no risk.For the genetically modified LAB, a formaldecision scheme has been developed.30

Two aspects should be analyzed: the gen-eral safety of the fermented food productmade with genetically modified LAB andthe probability of transfer of genetic infor-mation from the LAB to other bacteria inthe GIT of consumers (Figure 10-10). Inthis figure, the following aspects related toa quantitative risk assessment are consid-ered: (1) probability of transfer of geneticmaterial from the GMO to other microor-ganisms in the GIT as a function of the resi-dence time of the GMO in the GIT, (2)whether the GMO remains alive, (3) whetherDNA from lysed GMOs are taken up by nor-mal inhabitants of the GIT, (4) the probabil-ity that transfer of genetic information of theGMO to normal inhabitants results in an eco-logical advantage of the transformed inhabit-ants and, if so, (5) whether that may pose ahealth risk to the consumer, or environmentalrisks. Unfortunately, insufficient data areavailable to carry out such risk assessment inthe proper way.

Brie StiltonCamembert RoquefortCoulommier Blue

Gorgonzola

Figure 10-9 Schematic outline of milk fermentation processes and their cheese products.

Internal mold-ripenedSurface ripened

CheddarEmmenthalParmesanPort du salut

BrickEdamGouda

LimburgNeufchatel

Pressed Not pressedPressed

Curd scalded Curd scalded Curd not scalded

Part. ferm. milk, treated with rennetFermented and treated with rennet

Milk

I (E). Genetically modified lactic acid bacterium (GMO)2(A). Determine the distribution of the residence time of GMO in the gastrointestinal tract (GIT) of consumers. Take the time

corresponding with 95% of this distribution curve as t(r).3(Q). Will the GMO lyse with P(b)> b in the GIT within t(r)?4(A). Determine the probability P(c) that intact cells of the GMO transfer genetic information to normal inhabitants of the GIT.

In these studies, use t(r) as contact time and the conditions of the GIT.5(Q). Is P(C)XX6(A). Determine the probability P(f) that DNA originating from lysed GMO transform normal inhabitants of the GIT (resulting

in transformed inhabitants).7(Q). Is P(f)>f?8(A). Determine whether the transformed inhabitants obtain an advantage over untransformed inhabitants in the GIT: A(i).

Define A(i) in either faster growth rates t'(g), better adhesion to epithelial cells h', or higher production of certain metabo-lites {p(x}', x= 1,....}.

9(A) Determine the probability P(e) that any of the events described under action 8 will result in the formation of a hazardous(transformed inhabitant produce toxin or that will replace beneficial microorganisms in the GIT) microorganism.

10(Q). Is {P(c) + P(f)}*P(e) > e?I1 (A). Determine also the probability P(d) that the GMO will be transformed by genetic material originating from the common

GIT microorganisms (=modified GMO).12(A). Determine whether the modified GMO gains an advantage over untransformed GMO: B(i). Define B(I) in either faster

growth rates t"(g), better adhesion to epithelial cells h", or higher production of certain metabolites {p(x)", x=1,....}.13(Q). \sP(d)>d?14(A). Determine the probability P(g) that any of the events described under action 12 will result in the formation of a hazard-

ous GMO.15(Q). \sP(d)*P(g)>g?16(E). This risk is unacceptable and the GMO should not be released.17(E). This risk is acceptable and the GMO can be released.A = action; E = entry or end; Q = question

Figure 10-10 A proposal for a structured assessment of the risk related to the introduction of genetically modi-fied lactic acid bacteria in (or as) food products.29

In fact, only fermented products ofwhich the safety has been demonstrated,ecological hazards have been proven to bezero, and consequently, the risk is zero,should be approved. However, assumingthat the ecological hazard is not zero, whichmeans that the newly introduced gene con-fers an ecological benefit on recipient cells,both risk assessments should result in anacceptably low risk. This can only beachieved if the probability of transfer of ge-netic information from the donor to recipi-ent cells is very low. For horizontal transferof genetic material between microorgan-isms, conjugation, transformation, andtransduction show a higher probability forplasmid DNA than for chromosomal DNA.Therefore, it is recommended that the for-eign DNA be integrated (preferably on apreknown locus) into the chromosome ofthe microorganism. Integration atpreknown loci therefore forms an impor-tant component of the risk cube presentedin Figure 10-3.

A parameter that is essential for properrisk assessment of events in the human GITis the residence time of genetically modi-fied microorganisms in the GIT, whetherthey lyse, whether they conjugate, orwhether the DNA liberated during lysis cantransform other cells, including humancells. Although the number of reliable datasets have increased during the last decade,they are still not sufficient to solve all of thequestions of the decision scheme given inFigure 10-10. Nevertheless, the data avail-able indicate that LAB can survive passagethrough the GIT.19 The functional microor-ganism in many cheese fermentations,Lactococcus lactis, survives up to threedays, although the survival rate is only 1-2%. Another noteworthy observation inthis study was that the PCR method coulddetect special DNA stretches of this bacte-rium for up to four days after ingestion,12

indicating that either the feces containednonviable L. lactis or that these stretches ofDNA were liberated during lysis of the bac-

terium. Experiments with phage M 13 DNAingested by mice showed that a small butmeasurable percentage of this DNAreached peripheral leukocytes, spleen, andliver via the intestinal wall mucosa andcould be covalently linked to mouseDNA.25 Even taking into account that theseexperiments are rather artificial, for aproper risk assessment, they have to beincluded.

From these studies, it can be concludedthat there is a probability that geneticallymodified microorganisms, including LAB,can transfer genetic information to other mi-croorganisms in the GIT. This information isanother strong argument to ban geneticallymodified microorganisms that contain an an-tibiotic resistance gene. In cases where anongenetically modified microorganism hassuch a gene, it should be deleted because fur-ther unnecessary spread of antibiotic-resis-tant genes creates very serious health and en-vironmental problems.

• Step 5 (Figure 10-5). The release of fecesof consumers into the environment is the fi-nal step to be assessed. If the newly intro-duced gene does not encode for a proteinthat provides an ecological advantage forthe recipient cells, then a risk assessment ofthis step is not necessary. If this is not thecase, then the decision scheme outlined inan earlier publication30 can be applied forthis step. However, this step may becomemore complex because it should take intoaccount whether the transfer of genetic in-formation from the LAB to recipient mi-croorganisms has occurred in the GIT ofconsumers and, if so, these modified recipi-ents should be evaluated as well.

In this example, LAB are the functional mi-croorganisms in the fermented product, but forother organisms that are generally recognized assafe (GRAS), such as Saccharomyces cerevi-siae, Kluyveromyces lactis, Penicillium roque-forti, P. camenberti, and A. niger, or A.awamori, the same hazard and risk assessmentcan be applied with similar results.

CONCLUSION

Fermented foods are considered by consumersas natural and healthy.7'8'14'22 The introduction ofgenetically modified plants, microorganisms, orenzymes into these products needs careful discus-sion with authorities and consumer organizations.Many fermented foods contain living microorgan-isms, and if these organisms are genetically modi-fied, one should consider them as a potentialsource of DNA for horizontal gene transfer be-cause a number of studies provide circumstantialevidence that, in rare cases, genes have been later-ally transmitted among Eubacteria, Archaea, andEukarya. However, extensive studies byPuhler's6 group showed that this transfer occursat extremely low frequency.

REFERENCES

1. Barra, D. & Simmaco, M. (1995). Amphibian skin: apromising resource for antimicrobial peptides. Tib Tech13,205-209.

2. Beuchat, L. R. (1983). Indigenous fermented foods. InBiotechnology. Vol. 5, pp. 477-528. Edited by H. -J. &G. Reed. Weinheim: Verlag Chemie.

3. Casteels, P. & Ampe, C. (1989). Apidaecins: antibacte-rial peptides from honeybees. Embo 8, 2387-2391.

4. Chervitz, S. A., Aravind, L., Sherlock, G., Ball, C. A.,Koonin, E. V., Dwight, S. S., Harris, M. A., Dolinski,K., Mohr, S., Smith, T., Weng, S., Cherry, J. M. &Botstein, D. (1998). Comparison of the complete proteinsets of worm and yeast: orthology and divergence. Sci-ence 282, 2022-2033.

5. Cornelissen, B. J. C., Hooft van Huijsduijnen, R. A. M.& BoI, J. F. (1986). A tobacco mosaic virus-induced to-bacco protein is homologous to the sweet-tasting proteinthaumatin. Nature 321, 531-532.

6. Droge, M., Punier, A. & Selbitschka, W. (1998). Hori-zontal gene transfer as biosafety issue: a natural phe-nomenon of public concern. J Biotech 64, 75-90.

7. Gasser, F. (1994). Safety of lactic acid bacteria and theiroccurrence in human clinical infections. Bull Inst Pas-teur 92, 45-667.

8. Gilliland, S. E. (1990). Health and nutrition benefits forlactic acid bacteria. FEMS Microbiol Rev 87, 175-188.

9. Hamstra, A. M. (1993). Consumer acceptance of foodbiotechnology. SWOKA Research Report 137, TheHague, The Netherlands.

In these discussions, the proven benefits to theconsumer should be explained in clear terms.The risk assessment should show that the safetyhazard related to the newly introduced gene(s) iszero, and consequently, the risk for consumer iszero. Whenever possible, without losing theconsumer benefit, the environmental hazardshould also be zero. If that is not possible, therisk assessment should show that the risk is ex-tremely low, preferably less than 10~8 per prod-uct unit produced. Consumer studies show thatprovided there is a clear benefit and risk is ab-sent, most consumers will accept these products.To gain the confidence of the consumer for fer-mented foods in which recombinant DNA tech-nology plays a role, clear and transparent label-ing and public relations will be required.

10. Hoffmann, T., GoIz, C. & Schieder , O. (1994). ForeignDNA sequences are received by a wild-type strain ofAspergillus niger after co-culture with transgenic higherplants. Curr Genet 27, 70-76.

11. Jorgensen, R. B. & Andersen, B. (1994). Spontaneoushybridisation between oil seed rape (Brassica napus)and weedy B. campestris: a risk of growing geneticallymodified oil seed rape. Am J Botany 81, 1620-1626.

12. Klijn, N., Weerkamp, A. H. & de Vos, W. M. (1995).Biosafety assessment of the application of geneticallymodified Lactococcus lactis spp. in the production offermented milk products. System Appl Microbiol 18,486^192.

13. Koonin, E. V., Tatusov, R. L. & Galperin, M. Y. (1998).Beyond complete genomes: from sequence to structureand function. Current Opinion in Structural Biology 8,355-363.

14. Lee, Y. -K. & Salminen, S. (1995). The coming of age ofprobiotics. Trends Foods Sd Technol 6, 241-245.

15. Lorenz, M. G. & Wackernagel, W. (1994). Bacterialgene transfer by natural gene transformation in the envi-ronment. Microbiol Rev 58, 563-602.

16. Mann, C. C. (1999). Crop scientists seek a new revolu-tion. Science 283, 310-314.

17. Mann, C. C. (1999). Genetic engineers aim to soup upcrop photosynthesis. Science 283, 314-316.

18. Miller, R. V. (1998, January). Bacterial gene swappingin nature. Sd Am 46-5 1 .

19. Pedrosa, M. C., Golner, B. B., Goldin, B. R., Barakat, S.,Dallal, G. & Russel, R. (1995). Survival of yoghurt-con-taining organisms and Lactobacillus gasseri (ADH) andtheir effect on bacterial enzyme activity in the gas-trointestinal tract of healthy and hypochlorhydric eld-erly subjects. Am J Clin Nutr 61, 353-358.

20. Ronald, P. C. (1997, November). Making rice diseaseresistant. SdAm 68-73.

21. Rowland, L, Wiseman, H., Sanders, T., Adlercreuz, H.& Bowey, E. (1999). Metabolism of oestrogens andphytooestrogens: role of the gut flora. Biochem SocTrans 27, 304-308.

22. Salminen, S., Bouley, C., Bouron-Ruault, M. -C.,Cummings, J. H., Franck, A., Gibson, G. R., Isolauri, E.,Moreau, M. -C., Roberfroid, M. & Rowland, I. (1998).Functional food science and gastrointestinal physiologyand function. Br J Nutr 80, S147-S171.

23. Salminen, S., Isolauri, E. & Salminen, E. (1996). Clini-cal uses of probiotics for stabilizing the gut mucosal bar-rier: successful strains and future challenges. Ant vanLeeuwenhoek 70, 347-358.

24. Schliiter, K., Futtere, J. & Potrykus, I. (1995). Horizon-tal gene transfer from a transgenic potato line to a bacte-rial pathogen (Erwinia chrysanthemi) occurs — if atall — at extremely low frequency. Biotechnol 13, 1094-1098.

25. Schubbert, R., Renz, D., Schmitz, B. & Doerfler, W.(1997). Foreign (M13) DNA ingested by mice reachesperipheral leukocytes, spleen, and liver via the intestinalwall mucosa and can be covalently linked to mouseDNA. Proc NatAcadSci USA 94, 961-966.

26. Stiles, M. E. (1996). Biopreservation by lactic acid bac-teria. Ant van Leeuwenhoek 70, 33 1-345.

27. Teuber, M. (1990). Production and use of chymosinfrom genetically altered microorganisms. LebensmittelIndMilchwirtschaftlS, 1118-1123.

28. Verrips, C. T. (1989). Growth of microorganisms incompartmentalized products. In Mechanisms of Actionof Food Preservation Procedures, pp. 363—393. Editedby G. W. Gould. London: Elsevier Applied Science.

29. Verrips, C. T. (1995). Structured risk assessment ofrDNA products and consumer acceptance of these prod-ucts. In Biotechnology. Vol. 12, pp. 157-196. Edited byH. -J. Rehm & G. Reed. Weinheim, Denmark: VCHVerlaggesellschaft mbH.

30. Verrips, C. T. & Van den Berg, D. J. C. (1996). Barriersto application of genetically modified lactic acid bacte-ria. Ant van Leeuwenhoek 70, 299-3 16.

31. Wells, J. M., Robinson, K., Chamberlain, L. M.,Schofield, K. M. & LePage, R. W. F. (1996). Lactic acidbacteria as vaccine delivery vehicles. Ant vanLeeuwenhoek 70, 317-330.

CHAPTER 11

Safety Assessment ofProbiotics and Starters

Seppo J. Salminen, Atte J. von Wright, Arthur C. Ouwehand,and Wilhelm H. Hozapfel

SAFETY ASPECTS FOR PROBIOTICSAND BIOTECHNOLOGY

Today, milk can be fermented into a widerange of different products with different fla-vors, consistencies, and structures. Non-dairyfermented products such as fermented veg-etables and cereals are increasingly popular indifferent parts of the world, and they put newdemands on starter cultures. Additionally, thefocus on probiotics has increased the demand onstarter culture properties. This poses great de-mands on the starter cultures and their character-ization, including their activity, stability, and re-sistance to bacteriophages. Specific species andstrains of starter cultures are used and selectedfor novel applications. Thus, it is also importantto consider the safety aspects of new and espe-cially novel starter cultures. Thus far, the selec-tion of general starter cultures has focused onimproved technological properties and more ef-fective natural strains, which are easily appliedfor production technology, industrial processes,and starter culture storage.

The selection of current and more effectiveprobiotic strains has focused on natural strainswith specific target properties such as adhesionto different target mucosal models, immuno-genic properties, and the production of antimi-crobial substances or competitive exclusion ofpathogens. Fighting pathogens with other mi-crobes has been recommended earlier.66 It islikely that in the future, more advanced biotech-nological selection and modification systems

will be applied. Thus, it is important to definethe safety assessment standards and assess thedifferences between natural and geneticallymodified strains of starters. Genetically modi-fied strains of lactobacilli or bifidobacteria havenot been introduced into the field of probioticsor starters yet. However, it is likely that they willappear in the future as a result of work presentlybeing conducted. In this respect, the EuropeanUnion has a suggested procedure and decisiontree for the safety assessment, as indicated byJonas and coworkers.42 This procedure should befollowed (see Chapter 10).

Probiotic products provide a variety of docu-mented health benefits to all consumers.4 Thereare no apparent safety concerns for immuno-competent subjects. However, today, one has toconsider the growing number of immunodefi-cient subjects. Specific probiotics used in foodshave been shown to block the adhesion of patho-genic bacteria to human intestinal mucosa and toprevent enteroinvasive pathogens from invadingthe mucosal cells by competitive exclusion, anti-microbial production, or cell aggregation.4'11

Other studies have provided good evidence ofstabilizing and normalizing the gut mucosal bar-rier, thus promoting the health of the host frominfants to the elderly.4

In the case of probiotic foods, the applicationof probiotics is similar to any fermented productcontaining organisms that are already present inthe human gastrointestinal tract (GIT). Thus, nospecific concerns appear in relation to the con-sumption of any fermented foods or foods con-

taining viable bacterial cultures, such asprobiotics. However, as disease-specific appli-cations increase, it may be necessary to furtherevaluate the safety aspects of novel probiotics.In addition, high concentrations of microorgan-isms that are well adapted to the human GIThave probably not been deliberately used in fer-mented foods before.

The use of starter cultures in milk products isbased on the development of lactic acid andother acids to decrease the pH resulting in mi-crobiologically stable and safe fermented prod-ucts. These principles have also been applied tomeat, vegetable, and cereal products. In broadterms, starter cultures enhance the safety ofthese products, but in the future, selection ofnovel strains will need safety assessment mecha-nisms when new species and even new generamay be used.

Safety Assessment of Probiotics and Starters

Four major approaches may be used to assessthe safety of probiotic and starter strains (Exhibit11-1). Several groups such as the EuropeanUnion Probiotic Demonstration Working Grouphave suggested safety assessment proce-dures.2'26'67'68 In general, the safety assessment ofmicrobial food supplements is not well devel-oped, and often a case-by-case approach has to

Exhibit 11-1 Approaches for Assessing the Safetyof Probiotic and Starter Strains

1 . Characterization of the genus, species, andstrain and its origin will provide an initialindication of the presumed safety in relationto known probiotic and starter strains

2. Studies on the intrinsic properties of eachspecific strain and its potential virulence fac-tors

3. Studies on adherence, invasion potential,and the pharmacokinetics of the strain

4. Studies into interactions between the strain,intestinal and mucosal microflora, and thehost

be taken. A review on past safety studies hasbeen presented by Donohue et at.19

Suggested Virulence Factors and OtherHarmful Properties

Several potential virulence factors have beenproposed for lactic acid bacteria (LAB) (Figure11-1). Many of these are known virulence fac-tors for "real" pathogens; their importance forgenerally non-pathogenic organisms such as lac-tobacilli is uncertain. No significant risk factorsor virulence factors have so far been identifiedfor LAB or bifidobacteria used as starters orprobiotics.

Strong adhesion has been suggested as a viru-lence factor because it may facilitate transloca-tion and platelet aggregation, and thus may beconnected with endocarditis.26'31'55 In a studycomparing clinical Lactobacillus isolates frombacteremia patients with current probiotic anddairy strains, no indication that clinical bacter-emia strains would have stronger adhesion thancommon probiotics was found.44 Clinical bacter-emia isolates also showed no general ability toaggregate platelets; in fact, most of them werenonaggregating strains. Also, the platelet-aggre-gating strains showed relatively low adhesion abil-ity in intestinal mucosal models. Thus, the relationof platelet aggregation and adhesion may be in-volved in some cases, but it does not appear to be astrict virulence factor for LAB. However, whenmore potent specific probiotic microbes are se-lected or modified from current probiotics, it maybe important to assess both platelet aggregationproperties and adhesion properties in more thanone model. The adhesion models may include theCaco-2 cell model simulating the intestinal epithe-lium and the fecal or intestinal mucus model tosimulate the mucosal barrier.43'69'76'77

The ability to adhere to extracellular matrix(ECM) components such as fibronectin, col-lagen, laminin, and so forth, has been suggestedto be of importance for pathogenic microorgan-isms. ECM components form the major constitu-ents of wounded tissues and can thus serve asreceptors for invading microorganisms. Interac-tion with these components is therefore consid-

ered to play a role in the earliest steps of bacte-rial pathogenesis.32 Such interaction has alsobeen suggested to promote passage to the circu-lation.31 For probiotics, it has also been sug-gested to be a beneficial property because theprobiotics may colonize the damaged tissue andprevent subsequent colonization by invasivepathogens.75 In fact, the healing of damaged gas-tric mucosa has been observed to be enhancedafter colonization by lactobacilli.22 Hemolysisand hemagglutination have been suggested aspotential virulence factors; hemolysis may leadto anemia and edema. The ability to causehemolysis, through the production of a-hemol-ysin, has been found to be general among Lacto-bacillus isolates from food, clinical, and fecalorigin,5'36 and may therefore not be a relevantproperty for potential virulence. Hemagglutina-tion, on the other hand, was found to be rare.5

The ability to form capsules has been sug-gested to play a role in pathogenesis because itmay protect from phagocytosis.72 This propertywas found to be rare among L. rhamnosus strainsfrom clinical, food, and fecal origin.5

Translocation, the passage of viable bacteriafrom the intestine to other sites in the body, islikely to be a normal event because viable intesti-nal organisms can be recovered from the mesen-teric lymph nodes of a small percentage of healthyanimals. LAB may also enter the circulationthrough other routes such as dental manipulation,complications from delivery in postpartumwomen, and patients with gastrointestinal com-plications.36 In the blood, bacteria will come incontact with the complement system that hasbacteriolytic activity. Resistance to comple-ment-mediated killing may enhance the survivalin the blood and thus increase virulence.59 Like-wise, resistance to phagocytosis may contributeto increased virulence.

Regarding the enterococci, several unre-solved issues deserve special attention. Pres-ently, it may be concluded that their virulence isstill not well understood, but adhesins, hemol-ysin, hyaluronidase, aggregation substance,gelatinase, and their ability to transfer antibioticresistance may be considered as putative viru-lence factors.24

Figure 11-1 Potential virulence factors suggested for lactic acid bacteria

Resistance to complement mediated killingResistance to phagocytosisHemagglutinationHemolysisPlatelet aggregation

Transferable antibioticresistance

Capsule

Enzymes:hyaluronidasegelatinaseDNAse, etc.

SAFETY OF CURRENT PROBIOTICS

The Role of the Host

Current probiotics have mostly been derivedfrom studies on the genera and species in thehealthy human GIT. Taxonomy forms the firstbasis for safety assessment. Lactobacilli andbifidobacteria have a natural association withthe infant and human mucosal surfaces of themouth, GIT, and genitourinary tract. This hasbeen considered as an indicator that they aresafe, common commensals of the healthy nor-mal human micro flora. Data available on currentprobiotics and starters attest to their safety infood use.

Lactobacillus species are rarely isolated frombacteremia cases. The incidence of Lactobacil-/ws-caused bacteremia has been estimated to be0.1-0.3%.26JO Surveillance studies have not re-ported any current probiotic or starter strains tobe involved in septicemia.70 However, two casestudies have been reported where probiotic or-ganisms may have been involved in the etiologyof disease — a liver abscess caused by L.rhamnosus indistinguishable from LactobacillusGG60 and a case of Lactobacillus endocarditispossibly caused by probiotic L. rhamnosus.50

The most likely source of the infection is oftenconsidered to be the intestine. Subjects withLactobacillus bacteremia usually have severeunderlying diseases, predisposing them to infec-tions. Factors considered to be predisposing toinfections of intestinal origin are diabetic state,poor nutritional status, compromised immunesystem (e.g., due to AIDS or transplantation),traumatic injury, and peritoneal inflammation.Dental manipulation has also been suggested tobe a potential risk factor.36'46'82

In immunocompromised hosts, currentprobiotics offer protective mechanisms includ-ing immune stimulation and colonization resis-tance to pathogens, which may give incentivesfor their use to protect immunodeficient patientsfrom pathogenic microbes. Studies with con-genitally immunodeficient mice, however,showed an increased mortality in newborn pupsto mothers that had been mono-associated withLactobacillus GG and L. reuteri.™ This indicates

that immunocompromised neonates may be asubpopulation deserving special attention, notonly for possible infection by LAB, but also byother microorganisms.

The presence of predisposing factors suggeststhat not only are bacterial properties important forthe development of Lactobacillus infection, butalso that the host properties play a significant role.

It is important to first verify potential viru-lence factors for probiotics in the immunodefi-cient host. Biotechnology may offer advantageshere for the development of disease-specifichealth microbes for the treatment of specific tar-get populations.81

Antibiotic resistance may be a property ofsome LAB, but it is generally not transferable, asindicated in studies with one probiotic.78 How-ever, the risk of transfer of antibiotic resistancemay be a significant problem for enterococci,other novel probiotics, and starter strains if thetransfer potential is not included in the selectionof novel strains (see Chapter 10).

Safety Assessment Procedures

In a European Demonstration project onprobiotic properties, the safety assessment ofprobiotic microbes has been a special target. Rec-ommendations for safety assessment procedureshave been made by the group, outlining the targetsand steps for safety assessment of both natural andnovel probiotics19'56'68 (Exhibit 1 1-2).

RELEVANCE OF A SOUNDTAXONOMIC BASIS

Strong competition in the probiotic market,on the one hand, and unresolved questionsaround the regulatory situation on the other, in-dicate the complexity of the present situationwith regard to newly selected functional strainsof organisms. Species level identification alonedoes not provide sufficient transparency re-quired by industry, responsible scientists, regu-latory bodies, and also the modern consumer. Itis a well-established fact that the species andeven genus designation may provide a strong in-dication of typical habitats and possible origin of

Exhibit 11-2 Suggested Steps for the Assessment ofthe Safety of Probiotic and Starter Strains67'68

1. Strains with a long history of safe use willbe safe as starters or probiotic strains.

2. Strains that belong to species for which nopathogenic strains are known but which donot have a history of safe use may be safe asprobiotic or starter strains, but should betreated as novel foods.

3. Strains that belong to species for whichpathogenic strains are known should betreated as novel foods.

4. Characterization of the intrinsic propertiesof new strains. Strains carrying transferableantibiotic resistance genes should not bemarketed.

5. Strains that are not properly taxonomicallydescribed (DNA-DNA hybridization,rRNA sequence determination) should notbe marketed.

6. Assessment of the acute and subacute toxic-ity of ingestion of extremely large amountsof the probiotic and starter strains should becarried out.

7. Estimation of the infective properties invitro, using human cell lines and intestinalmucus, and in animal models, using suble-thally irradiated or immunocompromisedanimals.

8. Assessment of the effects of the strains onthe composition and activity of the humanintestinal microflora

9. Epidemiological and postmarketing sur-veillance

an organism (Table 1 1-1). In addition, the gen-erally accepted safety and technical applicabilityof a species and especially of "new" strains maystrongly be indicated by the species or even ge-nus to which it belongs (Table 1 1-2). Particu-larly, strains selected for their specific functionalproperties have to be clearly characterized be-low the species level. This is in the interest of themanufacturer, after considerable investment inscreening and selection procedures and in clini-cal studies, and also for postmarketing surveil-lance. Molecular fingerprinting methods may beapplied as reliable and highly discriminatory

tools in order to meet these challenges. In thisway, requirements of legislative bodies and gov-ernmental control organs for exact informationon physiological features and nomenclature of"new" strains may be satisfied.

The long tradition of "safe and acceptable"application of traditional starter cultures (e.g.,for fermented milk products) does not exist formost probiotic strains on the present-day market.The longest history of proven health benefits and"safe use" is probably exemplified by the L.paracasei strain "Shirota" and some strains of theso-called L acidophilus "group." Presently,strains of these species, that is, the L acidophilusgroup (thereby including the "related" speciessuch as L. crispatus and L. johnsonii) (see Table11-1), the L. casei/paracasei "group," and alsosome Bifidobacterium spp., predominate inprobiotic yogurts.34'45'71 Most of these strains origi-nate from the human GIT and have confirmedmuch of the pioneering work of Reuter and co-workers in the 1960s48'61-63 on lactobacilli au-tochthonous to the human GIT. The "biotypes"suggested by Lerche & Reuter48 for differentiat-ing within the L acidophilus group have largelybeen supported by the DNA homology groupsreported by other workers in 1980.41'47 This fi-nally resulted in six phylogenetically relatedspecies presently considered to belong to this"group." Although independent species, theycannot be readily distinguished by phenotypicfeatures.34'51'71 Also in confirmation of earlierwork by Reuter et al, representatives of two fur-ther homofermentative groups, L. salivarius andthe L. casei group, involving strains ofL. paracasei and L. rhamnosus, are presentlyused in probiotic products. Among the hetero fer-mentative lactobacilli also shown by Reuter 'sgroup to be part of the normal microbial popula-tion of the human GIT, especially strains pres-ently classified as L reuteri, show potential asboth human and animal probiotics.84

As with the L. acidophilus group, confusionalso exists concerning the correct nomenclatureof the L. casei group. Following the suggestionof Collins et al.,n most L. casei strains weretransferred to the new species L. paracasei,while the former L. casei ssp. rhamnosus was

Table 11-1 Typical Properties of Species of the So-called "Acidophilus Group" (modified ace. to ref.35, 51 and Reuter (1997): personal communication).

mol%G + Cin "Biotypes" DNAHomologySpecies Habitat* in the DNA ace. to groups ace. to

Reference (48) (47) (41)

L acidophilus Most? 32-37 l,/ll Ia A-1Lamylovorus P/C 40 IV(III) Ib A-3L crispatus H/G 35-38 III Ic A-2L gallinarum G 33-36 — Id A-4Lgasseri H/C 33-35 I Il a B-1L.johnsonii H/P/G 32-38 I, Il lib B-2

*H = humans; P = pigs; C = cattle; G = poultry.

elevated to species status. Dicks et al.,ls how-ever, suggested the reclassification of L. easelATCC 393 (one of the few remaining strains inthis species) and L. rhamnosus ATCC 15820 asL. zeae, and the designation of strain ATCC 334as the neotype strain of L. casei, simultaneouslywith the rejection of the name L. paracasei. Sub-sequently, this suggestion was strongly sup-ported by phylogenetic data from Mori et al.53

and phenotypic data from Klein et al.45 Yet, be-

cause of nomenclature rules, the full suggestionof Dicks et a/.18 has not yet been accepted by thejudicial commission. Therefore, practically allstrains presently designated as L. casei are offi-cially in fact representatives of L. paracasei.This has been shown by a recent investigation onstrains from novel-type yogurts on the market.71

Functional properties and safety of particularstrains of L. casei, L. rhamnosus, L. acidophilus,L crispatus, L. gasseri, and L. johnsonii are in-

Table 11-2 General Classification of Starter Cultures in Potential Risk Groups

"Safe" Doubtful Risk Group

Bifidobacterium Bif. dentium Peptostreptococcus spp.Carnobacterium (most spp.) Carnobacterium piscicola (?)* Streptococcus spp. (most)Enterococcus casseliflavus Enterococus spp. (most)tEnt. malodoratus Lb. catenaformeEnt. mundtii Lb. rhamnosustEnt. saccharolyticus Lc. garvieaeLactobacillus (most spp.) VagococcusLactococcusLeuconostocOenococcusPediococcusStreptococcus thermophilusWeissella

*Some strains are pathogenic to fish.tSome strains of Ent. faecium and Ent. faecalis with a history of safe use are known.^Several strains of Lb. rhamnosus with a "safe history" are known.

creasingly being studied. Species such asL. rhamnosus, Enterococcus faecium, andE. faecalis are presently classified as "risk group2" organisms (i.e., potential pathogens) in Ger-many.7 Some strains, however, find application inprobiotic products, and are considered not to con-stitute any health risk because of their history ofsafe use (Table 1 1-2). Differentiation of thesestrains from those of, for example, clinical, envi-ronmental, or animal origin, poses a special chal-lenge that may be solved by modern moleculartyping techniques. These genotyping techniquesare valuable tools for the identification of LAB,either on species level or for differentiation ofstrains to the clonal level.35 Advantages includetheir universal applicability, whereas the high dis-criminatory power of particular techniques such aspulsed-field gel electrophoresis (PFGE) may en-able differentiation among closely related strainsof a similar phenotype. Depending on the particu-lar situation, molecular methods such as restric-tion enzyme analysis (REA), PFGE, randomlyamplified polymorphic DNA (RAPD-PCR), andribotyping may be applied on chromosomal DNAfor the typing of probiotic strains. Plasmid profil-ing, by contrast, is not considered suitable for thetyping of individual strains due to the instability ofextrachromosomal DNA.

REA may be applied for differentiationamong strains of L. acidophilus,64 L. casei/L.rhamnosus,3 and L. reuteri.3'13 As for other meth-ods relying on the separation of DNA fragmentsby electrophoresis, the complexity of the band-ing patterns necessitates the use of computer-aided multivariate analysis.9 In addition, the se-lection of the appropriate restriction enzyme(s)is essential in order to obtain revealing discrimi-natory patterns. PFGE has been used success-fully to differentiate strains of importantprobiotic bacteria such as bifidobacteria,65

L. casei,23 and L. acidophilus.64 Enabling differ-entiation among different clones of a particularspecies, PFGE was shown, for example, to dis-tinguish among individual strains of E. faecalison the basis of 25 patterns obtained, as comparedto only 7 patterns obtained with ribotyping.28

PFGE may therefore enable intraspecies differ-entiation between probiotic and clinical strains

of enterococci.57'58 Ribotyping, on the otherhand, is mainly applicable on species level, andreveals fingerprinting patterns that are morestable and more easily interpretable than thoseobtained by REA.9 In addition, this technique ishighly reproducible and allows the possible useof a universal probe for all species because of thesimilarity of conserved regions of ribosomalgenes.29'58 Whereas it has been found appropriatefor studying the diversity of strains of L reuteriand L. fermentum from the mouse ileum,54

ribotyping may also be applied to characterizestrains of different Lactobacillus species,54'64

particularly of the L. acidophilus group.64 In asimilar fashion, soluble cell protein finger-printings revealed by polyacrylamide gel-elec-trophoresis also enable species-level distinctionwithin the L. acidophilus group.45'58

The RAPD-PCR technique often revealspoorly reproducible patterns and, similar toPFGE, it needs to be performed under carefullycontrolled conditions. This technique has, how-ever, been used for species-level distinctionamong strains of Bifidobacterium65 and of the Lacidophilus group.21 Using three different arbi-trary primers, RAPD-PCR produced revealing in-formation concerning the genetic diversity amongdairy lactococcal strains.15 Applying a combina-tion of two decamer primers in a single PCR reac-tion, a multiplex RAPD-PCR technique enableddifferentiation of Lactobacillus strains from theGIT of mice.15 Oligonucleotide probes comple-mentary to rRNA gene targets may be used for insitu detection of strains in mixed populations ofpotentially probiotic lactobacilli,6'30'33'58 entero-cocci,6 and bifidobacteria.25

Amplified fragment length polymorphism(AFLP), another more recent fingerprintingtechnique, uses the selective amplification of re-striction fragments from a digest of total ge-nomic DNA. It is a powerful discriminatory fin-gerprinting technique, and, with the exception ofListeria monocytogenes,1 it has thus far been ap-plied mainly for species-level differentiation ofGram-negative species such as Xanthomonasspp.,39 Aeromonas spp.3S>39Acinetobacter spp.,40

Pseudomonas spp.,27 and Campylobacterstrains20 from poultry. It also provided a sound

basis for classification of Alcaligenes faecalis-like isolates as a new species, Ralstoniagilardii.12

TRANSFER OF ANTIBIOTICRESISTANCE

The roles of different physiological gene trans-fer mechanisms — transduction, conjugation, andtransformation — in Lactobacillus, Lactococcus,Pediococcus, and Streptococcus thermophilushave been reviewed recently.79 In transduction,DNA is conveyed from one host to another by themediation of an infecting bacteriophage; in conju-gation, the DNA transfer occurs by direct cell-to-cell contact; and in transformation, the DNA isconveyed through direct uptake of free solubleDNA by the organism. In principle, transduction islimited by the phage host range and is thus con-fined within bacterial strains that are very closelyrelated to each other. In conjugation and transfor-mation, the DNA exchange can occur across spe-cies and even genera. Although both transductionand conjugation seem to be relatively commonphenomena among LAB, there has been little evi-dence of physiological transformation in thesegenera and species. Conjugation, often accompa-nied by the production of pheromone type of pro-teins, is also common among enterococci.10

With the notable exception of the enterococci,antibiotic resistance plasmids and transposablegenetic elements are relatively rare amongLAB,79 and apparently no antibiotic transferableresistance determinants have, so far, been de-tected among the bifidobacteria.49 Among thelactobacilli, certain strains of L. fermentum,L.acidophilus, L. reuteri, and L. plantarum havebeen shown to harbor resistance plasmidsagainst erythromycin (and related macrolides),tetracycline, and chloramphenicol.68 Thus, thepossibility of antibiotic resistance spreadingfrom the LAB common in fermented foodsseems to be remote. However, the prevalence ofantibiotic resistance in enterococcal strains usedor prevalent in many indigenous fermentedfoods needs to be evaluated.

From a safety point of view, the crucial ques-tion is whether antibiotic resistance determi-

nants conveying resistance to clinically impor-tant drugs could be transferred in the intestinefrom harmless commensals to actual or opportu-nistic pathogens. Although experimental evi-dence from real-life situations either in humansor in conventionally reared animals is difficult toproduce, some data obtained by using gnotobi-otic animals indicate the presumable conjuga-tional in vivo transfer of antibiotic resistancemarkers. For example, the transfer of an entero-coccal wide host-range plasmid pAMBl hasbeen observed between different genera of LABin the gut of germ-free mice.14'52 Thus, the spreadof antibiotic resistance among the intestinal mi-croflora remains a possibility, and care shouldbe taken in choosing strains, either for conven-tional fermented foods or for probiotic purposes,to avoid strains carrying transferable drug resis-tance determinants and strains that have the abil-ity to facilitate plasmid transfer.

When evaluating the antibiotic resistance pro-files among different species and strains, it iscrucial to differentiate between intrinsic resis-tance and resistance that is mediated by specialgenetic elements. Vancomycin resistance is aparticularly illustrative example. Vancomycin isone of the last resort antibiotics against multire-sistant staphylococci. Therefore, transferablevancomycin resistance among the enterococcalstrains is a serious clinical problem, both be-cause of the increasing incidence of enterococ-cal nosocomial infections and also in view of thepossible transfer of the resistance to staphylo-coccal strains.4 By contrast, the intrinsic resis-tance in several genera and species of LAB (i.e.,L. rhamnosus) due to the special structure of thecell wall does not pose any risk.8'30'78 These bac-teria are not pathogens, they do not transfer theresistance to other species or strains, and in therare occasions where they might be involved inopportunistic infections, there are plenty of ef-fective alternative antibiotics to use.

GENETICALLY MODIFIEDORGANISMS

Questions concerning the safety of geneti-cally modified organisms (GMOs) have increas-

ing relevance for LAB because recombinantDNA techniques now find general application tomany of these species and strains.79 Although nogenetically modified probiotics or starters areyet on the market, extensive research is concen-trated on the development of LAB-based oralvaccines.74'83 This application is very close to theprobiotic concept and should it become legallyacceptable and gain the confidence of the public,it is conceivable that recombinant-DNA tech-niques could also be used to improve the health-promoting or technological properties ofhuman or animal probiotic strains. Geneticmodification may also improve the properties ofstarter strains (e.g., by shortening the ripeningprocess or through increased production ofaroma compounds).

When food-associated microorganisms are tobe genetically modified, the safety of the result-ing GMO should be exceptionally well guaran-teed. The current EU-legislation classifies foodscontaining GMOs automatically as novel foodsand requires a rigorous safety evaluation accord-ing to the multistep procedure defined in the EUNovel Food Regulation.37 In the United States,the Food and Drug Administration does notmake a rigorous distinction between traditionaland genetically modified foods, expecting simi-lar safety standards of both and requiring a spe-cific approval of the latter only in cases wherethey differ significantly from their conventionalcounterparts.16

It is generally considered unacceptable thatfood-associated GMOs contain antibiotic resis-tance markers in the final genetic constructs.Different food-grade selection systems havebeen developed for the genetic modification ofLAB, mainly lactococci. These can be based onsugar fermentation genes, components of thelactic acid bacterial proteolytic systems, andbacteriocin resistance.17'79 Although these mark-ers are often more difficult to apply than the tra-ditional antibiotic resistance used in the geneticstudies, they are being constantly developed. Itis therefore to be expected that similar constructswill also be available for the lactobacilli.

In cases where the genetic modification ofstarter or probiotic strains is limited to the trans-

fer of well-known traits from strains of the sameor closely related species, the procedure can beregarded as self-cloning and hardly entails anyrisks greater than those already connected withthe use of unmodified parental strains. If, how-ever, genetic constructs containing genes fromcompletely different organisms are used, and theresulting GMOs have properties that are notnaturally occurring among strains of the species,the safety of the strain has to be carefully consid-ered, also taking into account the environmentalconsequences, in addition to the well-being ofthe host. The latter might become important ifthe genetic modification intentionally or adven-titiously could alter or expand the ecologicalniche of the organism. From the point of view ofthe host, the harmlessness of new gene productshas to be established in terms of the amountslikely to be present in foods or feeds or in theintestine. The safety depends naturally on thephysiological activity and fate of the substancesin question.

At present, the above considerations are hypo-thetical because no such novel starter orprobiotic GMOs have been introduced to mar-kets yet. Knowing the rapid developments in thisfield, however, the situation may change at anymoment. The enterprises that are first to marketgenetically modified starters or probiotics will,undoubtedly, create the administrative prece-dent cases that will be critical for future evalua-tion and approval procedures.

CONCLUSION

Foods containing LAB have an important nu-tritional value and foods containing probioticLAB may also have direct health effects. In mostcases, these microorganisms will be ingestedalive, and the assurance of safety of the con-sumer is therefore of major importance.

Starter strains often have a long history of safeuse. For probiotics, this is often not the case. Ingeneral, lactobacilli are rarely associated withdisease; thus far, only two cases are knownwhere probiotic lactobacilli may have been in-volved in disease. However, potential risk

groups can be identified, mainly immunocom-promised subjects. No common virulence fac-tors have so far been identified among lactoba-cilli; the main potential risk factor so far isantibiotic resistance and the ability to transferthis trait. This is also one of the main reasons

REFERENCES

1. Aarts, H. J. M., Hakemulder, L. E. & Van Hoef, A. M.A. (1999). Genomic typing of Listeria monocytogenesstrains by automated laser fluorescnce analysis of ampli-fied fragment length polymorphism fingerprinting pat-terns. Int J Food Microbiol 49, 95-102.

2. Adams, M. R. & Marteau, P. (1995). On the safety oflactic acid bacteria from food. Int J Food Microbiol 27,263-264.

3. Ahrne, S. & Molin, G. (1997). Restriction endonucleaseanalysis of total chromosomal DNA of Lactobacillus.Microecol Ther 26, 27-30.

4. Arthur, M., Reynolds, P. & Courvalin, P. (1996). Glyco-peptide resistance in enterococci. Trends Microbiol 4,401^407.

5. Baumgartner, A., Kueffer, M., Simmen, A. & Grand, M.(1998). Relatedness of Lactobacillus rhamnosus strainsisolated from clinical specimens and such from food-stuffs, humans and technology. Lebensm Wiss Technol31,489-494.

6. Betzl, D., Ludwig, W. & Schleifer, K. H. (1990). Identi-fication of lactococci and enterococci by colonyhybridisation with 23 S rRNA-targeted oligonucleotideprobes. Appl Environ Microbiol 56, 2927-2929.

7. BG Chemie (Berufsgenossenschaft der chemischenIndustrie). Sichere Biotechnologie. Eingruppierungbiologischer Agenzien: Bakterien. Merkblatt B 006, 12/95; ZH 1/346. Jedermann-Verlag Dr. Otto Pfeffer oHG,Heidelberg.

8. Billot-Kelin, D., Gutman, L., Sable, S., Guittet, E. &Van Heijenoort, J. (1994). Modification of Peptidogly-can precursors is a common feature of the low-levelVancomycin resistant species of Lactobacillus casei,Pediococcus pentosaceus, Leuconostoc mes enter oides,and Enterococcus gallinarum. J Bacteriol 176, 2398—2405.

9. Charteris, W. P., Kelly, P. M., Morelli, L. & Collins, J.K. (1997). Selective detection, enumeration and identi-fication of potentially probiotic Lactobacillus andBifidobacterium species in mixed bacterial populations.Int J Food Microbiol?) 5, 1-27.

10. Clewell, D. B. (1981). Plasmids, drug resistance, andgene transfer in the genus Streptococcus. Microbiol Rev45, 409-436.

why care should be taken when enterococci areto be used as probiotics or starter cultures.

In conclusion, there is no indication that thegeneral public is at risk from the consumption oflactobacilli or bifidobacteria used as probioticsor starters.

11. Cocconier, M. H., Bernet, M. F., Chauviere, G. &Servin, A. L. (1993). Adhering heat-killed human Lac-tobacillus acidophilus, strain LB, inhibits the process ofpathogenicity of diarrhoeagenic bacteria in cultured hu-man intestinal cells. J Diarrhoeal Dis Res 11, 235-242.

12. Coenye, T., Falsen, E., Vancanneyt, M., Hoste, B.,Govan, J. R. W., Kersters, K. & Vandamme, P. (1999).Classification of Alcaligenes faecalis-like isolates fromthe environment and human clinical samples asRalstonia gilardii sp. nov. Int J Syst Bacteriol 49, 405-413.

13. Collins, M. D., Ash, C., Farrow, J. A. E., Wallbanks, S.& Williams, A. M. (1989). 16S ribosomal ribonucleicacid sequence analysis of lactococci and related taxa.Description of Vagococcus fluvialis gen. nov., sp. nov. JAppl Bacteriol 67, 453-460.

14. Connel, M. A., Mercer, A. A. & Tannock, G. W. (1991).Transfer of plasmid pAMfll between members of thenormal microflora inhabiting the murine digestive tractand modification of the plasmid in a Lactobacillusreuteri host. Microb Ecol Health Dis 4, 343-355.

15. Daud Khaled, A. K., Neilan, B. A., Henriksson, A. &Conway, P. L. (1997). Identification and phylogeneticanalysis of Lactobacillus using multiplex RAPD-PCR.FEMS Microbiol Lett 153, 191-197.

16. Department of Health and Human Services, Food andDrug Administration. (1992). Statement of policy: foodsderived from new plant varieties. Federal Register57(104).

17. De Vos, W. M. (1999). Safe and sustainable systems forfood-grade fermentations by genetically modified lacticacid bacteria. Int Dairy J 9, 3-10.

18. Dicks, L. M. T., Du Plessis, E. M., Dellaglio, F. &Lauer, E. (1996). Reclassification of Lactobacillus caseiATCC 3493 and Lactobacillus rhamnosus ATCC 15820as Lactobacillus zeae nom. rev., designation of ATCC334 as neotype of L. casei subsp. casei, and rejection ofthe name Lactobacillus paracasei. Int J Syst Bacteriol46, 337-340.

19. Donohue, D., Salminen, S. & Marteau, P. (1998). Safetyof probiotic bacteria. In Lactic Acid Bacteria: Microbi-ology and Functional Aspects, pp. 369-384. Edited byS. Salminen & A. Von Wright. New York: MarcelDekker.

20. Duim, B., Wassenaar, T. M., Rigter, A. & Wagenaar, J.(1999). High-resolution genotyping of Campylobacterstrains isolated from poultry and humans with amplifiedfragment length polymorphism fingerprinting. ApplEnviron Microbiol 65, 2369-2375.

21. Du Plessis, E. M. & Dicks, L. M. T. (1995). Evaluationof random amplified polymorphic DNA (RAPD)-PCRas a method to differentiate Lactobacillus acidophilus,Lactobacillus crispatus, Lactobacillus amylovorus,Lactobacillus gallinarum, Lactobacillus gasseri, andLactobacillus johnsonii. Curr Microbiol 31,1 14—1 18.

22. Elliott, S. E., Buret, A., McKnight, W., Miller, M. J. S.& Wallace, J. L. (1998). Bacteria rapidly colonize andmodulate healing of gastric ulcers in rats. Am J Physiol275, G425-G432.

23. Ferrero, M., Cesena, C., Morelli, L., Scolari, G. &Vescovo, M. (1996). Molecular characterisation of Lac-tobacillus casei strains. FEMS Microbiol Lett 140, 215-219.

24. Franz, C. M. A. P., Holzapfel, W. H. & Stiles, M. E.(1999). Enterococci at the crossroad of food safety? Int JFood Microbiol 47 ', 1-24.

25. Frothingham, R., Duncan, A. J. & Wilson, K. H. (1993).Ribosomal DNA sequences of bifidobacteria: implica-tions for sequence-based identification of the human co-Ionic flora. Microbiol Ecol Health Dis 6, 23-27.

26. Gasser, F. (1994). Safety of lactic acid bacteria and theiroccurrence in human clinical infections. Bull Inst Pas-teur 92, 45-67.

27. Geornaras, I., Kunene, N. F., Von Holy, A. & Hastings,J. W. (1999). Amplified fragment length polymorphismfingerprinting of Pseudomonas strains from a poultryprocessing plant. Appl Environ Microbiol 65, 3828-3833.

28. Gordillo, M. E., Singh, K. V. & Murray, B. E. (1993).Comparison of ribotyping and pulsed- field gel electro-phoresis for subspecies differentiation of strains of En-terococcusfaecalis. J Clin Microbiol 31, 1570-1574.

29. Grimont, F. & Grimont, P. D. A. (1986). Ribosomal ri-bonucleic acid gene restriction as potential taxonomictools. Ann Inst Pasteur /Microbiol 137B, 165-175.

30. Handwerger, S., Pucci, M. J., VoIk, J. K., Liu, J. & Lee,M. S. (1994). Vancomycin-resistant Leuconostocmesenteroides and Lactobacillus casei synthesize cyto-plasmic peptidoglycan precursors that terminate in lac-tate. JBacteriol 176, 260-264.

31. Harty, D. W. S., Oakey, H. J., Patrikakis, M., Hume, B.B. H. & Knox, K.W. (1994). Pathogenic potential of lac-tobacilli. Int J Food Microbiol 24, 179-189.

32. Hasty, D. L., Courtney, H. S., Sokurenko, E. V. & Ofek,I. (1994). Bacteria-extracellular matrix interactions. InFimbriae: Adhesion, Genetics, Biogenesis, and Vac-cines, pp. 197-211. Edited by P. Klemm. Boca Raton,FL: CRC Press.

33. Hertel, C., Ludwig, W., Pot, B., Kersters, K. &Schleifer, K. H. (1993). Differentiation of lactobacillioccurring in fermented milk products by using oligo-nucleotide probes and electrophoresis protein profiles.Syst Appl Microbiol 16, 463-467.

34. Holzapfel, W. H., Haberer, P., Snel, J., Schillinger, U. &Huis in't Veld, J. H. J. (1998). Overview of gut flora andprobiotics. Int J Food Microbiol 41, 85-101.

35. Holzapfel, W. H., Haberer, P., Geisen, R., Bjorkroth, J.& Schillinger, U. (2000). Taxonomy and important fea-tures of probiotic microorganisms in food and nutrition.Am J Clin Nutr (in press).

36. Horwitch, C. A., Furseth, H. A., Larson, A. M., Jones, T.L., Olliffe, J. F. & Spach, D. H. (1995). Lactobacillemiain three patients with aids. Clin Infect Dis 21, 1460-1462.

37. Hugget, A. C. & Conzelmann, C. (1997). EU regulationon novel foods: consequences for the food industry.Food Sd Technol 8, 133-139.

38. Huys, G., Coopman, R., Janssen, P. & Kersters, K.(1996). High-resolution genotypic analysis of the genusAeromonas by AFLP-fingerprinting. Int J Syst Bacteriol46, 572-580.

39. Janssen, P., Coopman, R., Huys, G., Swings, J., Bleeker,M., Vos, P., Zaubeau, M. & Kerstens, K. (1996). Evalu-ation of the DNA fingerprinting method AFLP as a newtool in bacterial taxonomy. Microbiology 142, 1881-1893.

40. Janssen, P., Maquelin, K., Coopman, R., Tjernberg, I.,Bouvet, P., Kersters, K. & Dijkshoorn, L. (1997). Dis-crimination of Acinetobacter genomic species by AFLPfingerprinting. Int J Syst Bacteriol 47, 1 1 79-1 1 87.

41 . Johnson, J. L., Phelps, C. F., Cummins, C. S., London, J.& Gasser, F. (1990). Taxonomy of the Lactobacillusacidophilus group. Int J Syst Bacteriol 30, 53-68.

42. Jonas, D. A., Antignac, E., Antoine, J. M., Classen, H.G., Huggett, A., Knudsen, I., Mahler, J., Ockhuizen, T.,Smith, M., Teuber, M., Walker, R. & De Vogel, D.(1996). The safety assessment of novel foods: guide-lines prepared by the ILSI Europe novel foods taskforce. Food Chem Toxicol 34, 931-940.

43. Kirjavainen, P. V., Ouwehand, A. C., Isolauri, E. &Salminen, S. J. (1998). The ability of probiotic bacteriato bind to human intestinal mucus. FEMS Microb Lett167, 185-189.

44. Kirjavainen, P. V., Tuomola, E. M., Crittenden, R. G.,Ouwehand, A. C., Harty, D. W. S., Morris, L. F.,Rautelin, H., Playne, M. J., Donohue, D. C. & Salminen,S. J. (1998). In vitro adhesion and platelet aggregationproperties of bacteremia-associated lactobacilli. InfectImmun 67, 2653-2655.

45. Klein, G., Pack, A., Bonaparte, C. & Reuter, G. (1998).Taxonomy and physiology of probiotic lactic acid bacte-ria. Int J Food Microbiol 41, 103-125.

46. Larvol, L., Monier, A., Besnier, P. & Levecq, H. (1996).Abces hepatique a Lactobacillus acidophilus.Gastroenterol Clin Biol 20, 193-195.

47. Lauer, E., Helming, C. & Kandler, O. (1980). Heteroge-neity of the species Lactobacillus acidophilus (Moro)Hansen and Moquot as revealed by biochemical charac-teristics and DNA-DNA hybridisation. ZbI Bakt Hyg IAbtOrigCl, 150-168.

48. Lerche, M. & Reuter, G. (1962). Das Vorkommen aerobwachsender gram-positiver Stabchen des Genus Lacto-bacillus Beijerinck im Darminhalt erwachsenerMenschen. ZblBaktIAbt Orig 185, 446-481.

49. Lim, K. S., Huh, C. S. & Back, Y. J. (1993). Antimicro-bial susceptibility of Bifidobacteria. J Dairy Sd 76,2168-2174.

50. Mackay, A. D., Taylor, M. B., Kibbler, C. C. &Hamilton-Miller, J. M. T. (1999). Lactobacillus en-docarditis caused by a probiotic organism. ClinMicrobiol Infect 5, 290-292.

51. Mitsuoka, T. (1992). Intestinal flora and ageing. NutrRev 50, 438-^46.

52. Morelli, L., Sdarra, P. G. & Bottazzi, V. (1988). In Vivotransfer of pAMfi 1 from Lactobacillus reuteri to Entero-coccus faecalis. J Appl Bacteriol 65, 371—375.

53. Mori, K., Yamazaki, K., Ishiyama, T., Katsumata, M.,Kobayashi, K., Kawai, Y., Inoue, N. & Shinano, H.(1997). Comparative sequence analysis of the genes en-coding for 16S rRNA of Lactobacillus casei-relatedtaxa. Int J Syst Bacteriol 47, 54-57.

54. Ning, W., Mackie, R. I. & Gaskins, H. R. (1997). Bio-type and ribotype diversity among Lactobacillus iso-lates from mouse ileum. Syst Appl Microbiol 20, 423-431.

55. Oakey, H. J., Harty, D. W. S. & Knox, K. W. (1995).Enzyme production by lactobacilli and the potential linkwith infective endocarditis. J Appl Bacteriol 78, 142-148.

56. Ouwehand, A. C. & Salminen, S. J. (1998). The healtheffects of viable and non-viable cultured milks. IntDairy J 8,749-758.

57. Piehl, I., Biilte, M. & Reuter, G. (1995). Pulsfeld-Gelelektroforese zur Feindifferenzierung vonMikroorganismen - ein Beitrag zur Charakterisierungvon Probiotika-Stammen. In Proc 36th Arbeitstagungdes Arbeitsgebietes "Lebensmittelhygiene, " Vol. 2. pp.226-233. GieBen: Deutsche VeterinarmedizinischeGesellschaft eV.

58. Pot, B., Hertel, C., Ludwig, W., Descheemaker, P.,Kersters, K. & Schleifer, K. H. (1993). Identificationand classification of Lactobacillus acidophilus, L.gasseri and L. johnsonii strains by SDS PAGE andrRNA-targeted oligonucleotide probe hybridisation. JGen Microbiol 139, 513-517.

59. Prescott, L. M., Harley, J. P. & Klein, D. A. (1993). Mi-crobiology. Dubuque, IA: Wm. C. Brown Publishers.

60. Rautio, M., Jousimies-Somer, H., Kauma, H.,Pietarinen, L, Saxelin, M., Tynkkynen, S. & Koskela,M. (1999). Liver abscess due to a Lactobacillusrhamnosus indistinguishable from L. rhamnosus strainGG. Clin Infect Dis 28, 1159-1160.

61. Reuter, G. (1965). Das Vorkommen von Laktobazillenin Lebensmitteln und ihr Verhalten im menschlichenIntestinaltrakt. Zentralbl Bakt Hyg I Abt Orig A 197,468^87.

62. Reuter, G. (1965). Untersuchungen iiber die Zusam-mensetzung und BeeinfluBbarkeit der menschlichenMagen- und Darmflora unter besonderer Beriick-sichtigung der Laktobazillen. Erndhrungsforsch 10,429^35.

63. Reuter, G. (1969). Zusammensetzung und Andwendungvon Bakterienkulturen fur therapeutische Zwecke.Arzneim-Forsch Drug Res 50, 951-954.

64. Roussel, Y., Colmin, C., Simonet, J. M. & Decaris, B.(1993). Strain characterisation, genome size and plas-mid content in the Lactobacillus acidophilus group(Hansen and Mocquot). J Appl Bacteriol 74, 549-556.

65. Roy, D., Ward, P. & Champagne, G. (1996). Differen-tiation of bifidobacteria by use of pulsed-field gel elec-trophoresis and polymerase chain reaction. Int J FoodMicrobiol 29,11-29.

66. Saavedra, J. M. (1995). Microbes to fight microbes: anot so novel approach to controlling disease. J PediatrGastroenterol Nutr 21,1 25-1 29.

67. Salminen, S. & Von Wright, A. (1998). Currentprobiotics-safety assured? Microb Ecol Health Dis 10,68-77.

68. Salminen, S., Von Wright, A., Morelli, L., Marteau, P.,Brassard, D., De Vos, W., Fonden, R., Saxelin, M.,Collins, K., Mogensen, G., Birkeland, S. -E. & Mattila-Sandholm, T. (1998). Demonstration of safety ofprobiotics: a review. Int J Food Microbiol 44, 93-106.

69. Salminen, S., Bouley, C., Boutron-Rualt, M. -C.,Cummings, J. H., Franck, A., Gibson, G., Isolauri, E.,Moreau, M. -C., Roberfroid, M. & Rowland, I. (1998).Functional food science and gastrointestinal physiologyand function. BrJNutrSuppl 1, 147-171.

70. Saxelin, M., Chuang, N. H., Chassy, B., Rautelin, H.,Makela, P. H., Salminen, S. & Gorbach, S. L. (1996).Lactobacilli and bacteremia in Southern Finland: 1989-1992. Clin Infect Dis 22, 564-566.

71. Schillinger, U. (1999). Isolation and identification oflactobacilli from novel-type probiotic and mild yoghurtsand their stability during refrigerated storage. Int J FoodMicrobiol 47, 79-87.

72. Sims, W. (1964). A pathogenic Lactobacillus. JPatholBacteriol 87, 99-105.

73. Stahl, M. & Molin, G. (1994). Classification of Lactoba-cillus reuteri by restriction endonuclease analysis ofchromosomal DNA. Int J Syst Bacteriol 44, 9-14.

74. Steidler, L., Robinson, K., Chamberlain, L., Schofield,K. M., Reault, E., Le Page, R. W. & Wells, J. M. (1998).Mucosal delivery of murine interleukin-2 (IL-2) and IL-6 by recombinant strains of Lactococcus lactiscoexpressing antigen and cytokine. Infect Immun 66,3183-3189.

75. Toba, T., Virkola, R., Westerlund, B., Bjorkman, Y.,Sillanpaa, J,. Vartio, T., Kalkkinen, N. & Korhonen, T.K. (1995). A collagen-binding S-layer protein in Lacto-bacillus crispatus. Appl Environ Microbiol 61, 2467-2471.

76. Tuomola, (nee Lehto) E. M. & Salminen, S. J. (1998).Adhesion of some probiotic and dairy Lactobacillusstrains to Caco-2 cell cultures. Int J Food Microbiol 41,45-51.

77. Tuomola, E., Ouwehand, A. C. & Salminen, S. (1999).Human ileostomy glycoproteins as a model for small in-testinal mucus to investigate adhesion of probiotics. LettAppl Microbiol 28, 159-163.

78. Tynkkynen, S., Singh, K. & Varmanen, P. (1998). Van-comycin resistance factor of Lactobacillus rhamnosusGG in relation to enterococcal vancomycin resistance(van} genes. Int J Food Microbiol 41,1 95-204.

79. Von Wright, A. & Sibakov, M. (1998). Genetic modifi-cation of lactic acid bacteria. In Lactic Acid Bacteria:Microbiology and Functional Aspects, pp. 161-210. Ed-ited by S. Salminen & A. Von Wright. New York:Marcel Dekker.

80. Wagner, R. D., Warner, T., Robers, L., Farmer, J. &Balish, E. (1997). Colonization of congenitally irnmu-nodeficient mice with probiotic bacteria. Infect Immun65,3345-3351.

81. Wagner, R. D. & Balish, E. (1998). Potential hazards ofprobiotic bacteria for immunodeficient patients. Bull IntPasteur 96, 165-170.

82. Wells, C. L., Maddaus, M. A. & Simmons, R. (1988).Proposed mechanisms for the translocation of intestinalbacteria. Rev Infect Dis 10, 958-979.

83. Wells, J. M., Wilson, P. W., Norton, P. M., Gasson, M.J. & Le Page, R. W. F. (1993). Lactococcus lactis: highlevel expression of tetanus toxin fragment C and protec-tion against lethal challenge. MoI Microbiol 8, 1115.

84. Wolf, B. W., Garleb, K. A., Ataya, D. G. & Casas, I. A.(1995). Safety and tolerance of Lactobacillus reuteri inhealthy adult male subjects. Microb Ecol Health Dis 8,41-50.

CHAPTER 12

Practical Applications:Prospects and Pitfalls

Yasmine Motarjemi, A. Asante, Martin R. Adams, and M. J. Robert Nout

INTRODUCTION

Fermentation is one of the oldest technologiesused for food preservation. Over the centuries, ithas evolved, been refined, and diversified untiltoday, when a large variety of foods are derivedfrom this technology, both in the industrializedand the developing countries, in households,small-scale food industries, and large commer-cial enterprises. Fermented foods form a majorpart of the human diet all over the world. Insome regions, mainly in African countries, fer-mentation plays an important role in the nutri-tion of infants and young children because it isused for the preparation of complementaryfoods. f

Advances in food science and technologyhave given rise to a wide range of new food tech-nologies. Nevertheless, fermentation has re-

Note: This chapter is partly based on the report of aJoint Food and Agriculture Organization(FAO)/WorldHealth Organization (WHO) Workshop on fermenta-tion as a household technology to improve food safetyheld in Pretoria, South Africa, 1 1-15 December 1995.18

fThe term complementary foods refers to any nutri-ent-containing foods, be they solids or liquids, otherthan breast milk, that are given to infants and youngchildren during the period of complementary feeding.This period of complementary feeding corresponds tothe period during which other foods are providedalong with breast milk.

Source: Portions of this chapter are adapted fromFermentation: Assessment and Research. Report of aJoint F AO /WHO Workshop on Fermentation as aHousehold Technology to Improve Food Safety,Pretoria, South Africa, December 11-15, 1995, ©1996, World Health Organization.

mained one of the most important food process-ing techniques throughout human history. Manybenefits are attributed to fermentation. It canpreserve food (i.e., increase shelf life), improvedigestibility, enrich food, and enhance taste andflavor. It is also an affordable technology, andthus is accessible to all populations. Further-more, fermentation has the potential to enhancefood safety by controlling a great number ofpathogens in foods. Thus, it makes an importantcontribution to human nutrition, particularly indeveloping countries, where economic problemsare a major barrier to ensuring food safety.

In general, fermented foods, particularly thoseproduced under controlled conditions, have agood record of safety and are implicated in out-breaks of diseases relatively infrequently. Thereare, however, more concerns with traditional andartisanal productions, where the application offer-mentation is based largely on experience andknowledge gained through trial and error by gen-erations of food producers and households. Suchan empirical approach presents a major pitfall.Depending on the process, ingredients, raw mate-rial, and environmental conditions, the processmay lead to unsafe products.

This chapter reviews the importance of fer-mentation from a public health and food safetypoint of view. Specific emphasis is put on prob-lems in developing countries where risks of foodcontamination and disease are greater, and thepotential contribution of fermentation for pro-cessing and storage of foods more important.The chapter examines the risks and benefits offermentation for human nutrition, with a particu-lar focus on the fermentation of complementary

foods, and looks into the prospects for promot-ing the technology and improving the safety andnutritional quality of the derived products. Italso presents two applications of the HazardAnalysis and Critical Control Point (HACCP)concept to fermented foods as examples.

BENEFITS OF LACTIC ACIDFERMENTATION

Depending on the organism used (i.e., molds,yeasts, bacteria), there are different types of fer-mentation processes. Fermentation mediated bylactic acid bacteria (LAB) is the process that ispresently of greatest interest to food safety andpublic health. Other types of fermentation arealso important to public health because they maycontribute to food security and overcoming mal-nutrition. The focus of this chapter is on lacticfermentation.

There is considerable evidence to show that lac-tic acid fermentation inhibits the growth, survival,and toxin production of a number of pathogenicbacteria1 (see also Chapter 2). The inhibitory ef-fect of lactic acid fermentation on pathogenic bac-teria as well as the spoilage organisms is due to therapid growth and acid production by LAB, theconsequent decrease in pH, and the formation ofother antimicrobial factors associated with LAB,such as bacteriocins, hydrogen peroxide, ethanol,and diacetyl.1'8 The extent to which pathogens areinhibited depends on the organism concerned, thetemperature, the amount of acid produced, andthe properties of the food (e.g., the buffering ca-pacity). For instance, in cereals and vegetableproducts that are weakly buffered, an efficientlactic acid fermentation will produce a pH of 4or less, at which the growth of bacterial patho-gens is inhibited and many bacteria die. This as-pect of fermentation, as seen in the followingparagraphs, is of great importance to publichealth, particularly in developing countrieswhere safe food storage by cold or hot holding isdifficult because of socioeconomic constraints.

Lactic acid fermentation has also been associ-ated with the reduction of certain naturally oc-curring toxins in plant foods, notably a reductionof cyanogenic glycosides that are present in cas-

sava. Cassava is a major staple in Africa and,due to the presence of these compounds, it hasbeen associated with several health problems.51

Although the degradation of the cyanogenic gly-cosides is mediated by endogenous plant en-zymes, microbial activities carried out during fer-mentation contribute to the detoxification processby softening the plant tissues (see Chapter 4).

Nutritional benefits have also been attributedto fermentation. Some fermentation processeslead to enhanced digestibility of carbohydratesas a result of degradation of oligosaccharidesand dietary fiber, particularly prevalent in foodsof plant origin. The use of amylase-rich flour(ARF) combined with a small amount of a lacticacid starter culture can increase the nutrient den-sity of starchy foods while keeping a semi-liquidconsistency. This makes the fermented food par-ticularly interesting as complementary food forinfants and young children. Fermentation alsoprovides an optimum pH for the activity ofphytase and the degradation of phytate. The deg-radation of phytate increases the amount ofsoluble iron several fold, thereby increasing itsbioavailability. Lactic fermentation also de-creases the tannin content of cereals and, in thisway, has a positive impact on mineral absorptionand the protein digestibility of cereals.

Over and above the influence on pathogenicbacteria and toxic compounds leading to im-proved food safety and quality, specific benefi-cial health effects such as prophylaxis againstEscherichia coli and other pathogens andhypocholesterolemic and anticarcinogenic ef-fects have also been attributed to some selectedstrains of LAB. A number of such organismswith claimed health-related properties are al-ready used in food production and are availablein many countries. The use of probiotics raisestwo fundamental questions: (1) Are the organ-isms used safe? and (2) Are the health claimsvalid? The former has been discussed in Chapter1 1 of this book. The latter is the subject of exten-sive research. A number of studies have shownsome specific health benefits with regard to lac-tose intolerance and gastrointestinal disorders.So far, there is no conclusive explanation for theunderlying molecular mechanisms. Also, there

are a number of claims that have not been sub-stantiated and need further investigation.13'22

Advances in genetic modification open newdoors to fermentation technology.34 Geneticmodification may be carried out on the substrate(e.g., plants) or on the microorganisms used forfermentation. Fermentation may also be used forthe production of enzymes with enhanced prop-erties and improved quality (see Chapter 10).These can have significant benefits in terms ofimproving the fermentation process and thequality and safety of the final product.

Genetic modification can be used in differentways to improve the fermentation process and toenhance the quality and safety of the final prod-uct. For instance, starter cultures may be im-proved with regard to their rate of growth, poten-tial for lactic acid production, salt tolerance,inability to metabolize organic acids, resistanceto bacteriophage, production of flavor com-pounds, and so on.6'8 Genetic modification mayalso be used to remove or reduce hazards in theraw material, as in the production of a variety ofcassava containing less cyanogenic glycosides.

As discussed in Chapter 10, the prospect ofusing genetic modifications for better control offermentation processes and to enhance the safetyand quality of derived products has two precon-ditions: (1) an adequate method for assessing thesafety and nutritional implications of these prod-ucts and (2) consumer acceptance. FAO, WHO,and other organizations have developed strate-gies to assess the safety of such foods,17'50 andfurther work in this area is planned. To gain con-sumer acceptance, genetic modification needs tobe considered in the broader perspectives of risksand benefits for the consumer, environment, so-cioeconomic implications (in both developing andindustrialized countries), and, in general, the needfor such technology. The prospect of using geneticmodification technology for improving fermenta-tion processes relies on an appropriate risk com-munication with consumers.

It should be mentioned that genetic modifica-tion can also occur spontaneously in nature andlead to new starter strains. Conventional meth-ods of screening and selection of new starterstrains with desirable properties have been used

for the production of products with enhancedquality.34

PITFALLS OF FERMENTATION

As simple as it may seem, fermentation is alsoa sensitive and complex technology. The safetyand quality of the final product depend on anumber of factors, which need to be carefullycontrolled. These factors are

• quality and safety of raw materials, includ-ing the initial level of contamination

• levels of environmental hygiene and sanita-tion

• quality and safety of starter culture• safety of metabolites• processing conditions and degree of acidity

achieved

In certain applications, particularly in the con-text of cottage industry or household applicationof fermentation, it may be difficult to controlthese factors adequately. Not infrequently, fer-mented products have been incriminated in in-fections or intoxications. In certain settings, in-adequate fermentation has been a major cause offood-borne intoxications. For instance, in China,2,861 cases of botulism (745 outbreaks) werereported in the year 1989, causing some 421deaths. The major implicated foods were home-made fermented bean products. Fermented fishand salmon eggs are a significant cause of botu-lism occurring among the native AlaskanInuit population in the North American conti-nent. Fermented Inuit products such as whitewhale meat, seal flippers, or salmon eggs lackfermentable carbohydrates and therefore do notundergo a sufficient pH reduction to preventtoxigenesis.20 Cheeses of different categorieshave also been implicated in outbreaks of vari-ous types of food-borne diseases.24 Examples ofother documented outbreaks of illness are listedin Table 12-1. It is likely that outbreaks of food-borne disease caused by fermented foods pro-duced under unhygienic conditions or withfaulty handling of the products occur more fre-quently than are officially reported. However,

Table 12-1 Examples of Food-Borne Disease Outbreaks Associated with Fermented Foods

Implicated Food

VegetablesPaste of soybeans

and wax gourdsSauerkraut

Milk productsCurd (yogurt)YogurtHazelnut yogurt (hazelnut

puree was contaminated)Sour milkYogurt

Meat productsSemi-dry sausagesPork (salami)Pork (labh, raw and nahm,

fermented)Salami stick

SalamiSausages (Lebanon bologna)

FishFish (seal flipper)Fish (beaver tails)Fish (salmon fish heads)Salmon eggs

CheesesSoft cheese (Mexican-style)Soft cheese

CheeseSoft cheeseGoat milk cheeseCheddar cheese

Mozzarella cheeseCheeseCheese (Brie, Camembert,

Coulommiers)Cheese (Brie, Camembert)Cheese (Brie, Camembert)White processed cheeseMexican-style soft cheeseMexican-style soft cheeseHand-pressed direct set cheeseCheese

Swiss cheese

Causative Agent

Clostridium butyricumHistamine

Campylobacter jejuniClostridium perfringens

Clostridium botulinumClostridium botulinumEscherichia coli O1 57

Escherichia coli O11 1 :NMSalmonella anatum

TrichinellaSalmonella typhimurium

Escherichia coli O1 57Salmonella typhimurium

Clostridium botulinumClostridium botulinumClostridium botulinumClostridium botulinum

Listeria monocytogensSalmonella berta

Salmonella enteritidisSalmonella dublinSalmonella paratyphiSalmonella heidelberg

Salmonella typhimuriumEscherichia coli O1 57

Escherichia co//O124:B17Escherichia co//O27:H20Clostridium botulinumStreptococcal pharyngitisStreptococcus zooepidemicusBrucella melitensisStaphylococcus aureusStaphylococcus aureus and

Shigella sonneiHistamine

Numberof Cases

6

160167

271116

2352

2785 (including

13 secondary cases)2326

178

15*

14282 (including

3 secondary cases=700

42273339t

(28000-36000 est.)32122

387 est.17027197163116

>50**6

References

3230

733

364135

104

26

121138

39403920

24,27

159

2814

193

44

2933751633

4043

*15 cases involved in 7 outbreaks between 1971-1984.t339 cases where reported in Colorado. From the attack rates noted (28-36%) and the amount of cheese presumably

consumed (2830 kg), it is estimated that between 28,000 and 36,000 persons were affected in total.**Bacillary dysentery, which affected numerous persons who had eaten various French cheeses purchased at a Paris airport in

1982, was reported by several Scandinavian countries.

weaknesses in the food-borne disease surveil-lance system, particularly in regard to householdfood preparation, do not provide for identifica-tion and reporting of these outbreaks.

Although LAB can inhibit the growth of cer-tain food-borne pathogens, care should be takento ensure that the raw material is of high hy-gienic quality and the risk of contamination fromthe environment is minimized. A number offood-borne hazards are not controlled by lacticacid fermentation and so present a serious threatto health if they persist in food. For instance,enterohemorrhagic E. coll has shown patterns ofacid resistance and may survive certain fermen-tation processes. Yogurt and fermented meathave been recognized as potential vehicles ofenterohemorrhagic E. coli infection. Food andwaterborne viruses, a relatively frequent causeof gastroenteritis, may survive high levels ofacidity. For instance, Simian rotavirus has beenshown to survive a high level of acidity during24 hours of storage in model fermented foods53

(see Chapter 8). There is little information on theeffect of fermentation on parasites, such asCryptosporidium, Giardia lamblia, and food-borne trematodes. The cysts or metacercariae ofthese organisms often show resistance to ad-verse conditions (see Chapter 9). Thus, the pos-sibility that they may survive fermentationshould not be excluded. Most toxins producedby algae, bacteria, and molds are also unaffectedby fermentation (see Chapter 5).

In view of this, it is important not to rely onlyon the fermentation step per se to eliminate orreduce hazards to safe levels. To ensure safety, itis important to design the process in such a wayas to (1) prevent contamination, (2) eliminate orreduce the hazards as much as possible, and (3)control the growth of pathogens. For this reason,fermentation steps frequently need to be com-bined with other process operations such assoaking, cooking, and so forth. Lactic acid fer-mentation also has a limited effect onantinutritional factors, such as protease inhibi-tors and lectins, and other process operationssuch as heat treatment should be used to destroythese antinutritional factors.

There are also some concerns with the organicacids that are produced during lactic acid fer-

mentation. Microbially produced lactic acid isusually a mixture of the optical isomers L(+) andD(-) lactic acid. D(-) lactic acid cannot be me-tabolized by humans. Excessive production andintake of D(-) lactic acid may result in acidosis, adisturbance of the acid-alkali balance in the blood.However, very little is known about the "toxicity"of D(-) lactic acid for malnourished or sick chil-dren, and little data are available on the D(-) lacticacid content of fermented foods prepared at thehousehold level. Further studies in this area areneeded to elucidate the role of D(-) lactic acid andmethods to control its production.1'2'45'54

Histamine poisoning is also a potential prob-lem with fermented foods. Although histaminepoisoning has been commonly associated withthe consumption of scrombroid-type fish such astuna and mackerel, certain fermented foods suchas wine, soy sauce, and, in particular, cheese,may also present a risk. Poisoning due to hista-mine or other biogenic amines is caused by anaccumulation of histamine or other type of bio-genic amines (e.g., tyramine, phenylethylamine)following metabolic activity of decarboxylase-positive LAB such as Lactobacillus buchneri.However, the problem can be controlled to alarge extent by observing good manufacturingpractice during production. For instance, duringthe manufacture of cheese, the control of factorssuch as hygienic quality of milk and temperatureof storage minimize the risk of histamine forma-tion42 (see Chapter 6).

IMPORTANCE OF FOODFERMENTATION IN PUBLIC HEALTH

Food-borne diseases are a major public healthproblem all over the world, in both industrial-ized and developing countries. The industrial-ized countries, benefiting from a higher standardof living, good water supply and sanitation, andtechnologies for processing and preservingfoods, have succeeded in combating many food-borne infections. This has resulted in a reductionof a great number of food-borne infections suchas typhoid fever, cholera, and shigellosis. Never-theless, food-borne diseases remain a wide-spread public health problem, and statistics inthese countries indicate that possibly up to 30%

of the population may suffer from a food-borneillness annually.31

It is the developing world that bears the brunt ofthe problem. Although statistics on the incidenceof food-borne diseases are not available, the highprevalence of diarrheal diseases in these parts ofthe world, particularly in infants and young chil-dren, is an indication of an underlying food safetyproblem. In 1997, 4,000 million cases of diarrheawere estimated to occur in the world.52 Approxi-mately 1,500 million episodes of diarrhea occurannually in children under the age of five, andmore than 1 .8 million children die as a result. Indi-rectly, diarrheal diseases kill many more childrenbecause they are one of the major underlying fac-tors of malnutrition. It is estimated that annually,some 13 million children under the age of five diefrom the associated effect of malnutrition.

The etiological agents responsible for food-borne diseases are broad and include bacteria,viruses, and parasites, all of which have beendiscussed here individually in their respectivechapters.

The sources of contamination of food are di-verse and include polluted water, night soil,dust, flies, domestic animals, dirty utensils, andfood handlers. Raw foods themselves may alsobe a source of contaminants because many foodsharbor pathogens or come from infected ani-mals. Moreover, during the preparation process,there is an added risk of cross-contamination.However, one of the major factors leading tofood contamination is time-temperature abuseduring food preparation and storage with the re-sult that pathogens survive, grow, and producetoxins (Figure 12-1).

Food handler (e.g.,contaminated hands)

Files and pests

Contaminatedhousehold water

Polluted environment(soil, dust)

Dirty pots andcooking utensils

Cross-contamination

Human and animal excreta

Infected food animal

Night soil

Irrigation andWastewater"

Domestic animals

FOOD

(Raw/Cooked)

Time-Temperature

Abuse

Contaminated Food

Figure 12-1 Sources of food contamination

So

urc

e o

f food c

on

tam

ina

tion

Surv

ival

an

dG

row

th

In addition to agents of diarrheal diseases,food may also be a vehicle for chemical hazards,whether naturally present or contaminating thefood as a result of poor agricultural practices orenvironmental pollution as dealt with in Chap-ters 4, 5, and 6. Depending on the dose, chemicalhazards may lead to acute intoxication or long-term health problems such as cancers and otherchronic diseases. Food may also containantinutritional factors such as enzyme inhibitors,phytates, lectins, and polyphenols, which inter-fere with the digestion, absorption, or other as-pects of metabolism of nutrients in foods.

Food processing technologies are applied fora variety of reasons. They may be applied to ren-der food more digestible or edible, to retain orenhance sensory quality, to increase shelf life, toimprove nutritional quality, and/or to renderfood safe. Food fermentation has proved to pos-sess many of these features; in this way, it is animportant food technology. However, its role inimproving the nutritional quality of foods, par-ticularly complementary foods, and preservingfoods and preventing growth of most pathogenicorganisms make this technology particularly im-portant from a public health point of view.

Fermentation is particularly important forfood safety and for the prevention of diarrhealdiseases in the developing regions. Although theapplication of basic rules of food hygiene canprevent a great proportion of diarrheal diseases,in the developing countries, the application ofthese rules is sometimes hampered by socioeco-nomic constraints such as inadequate supply ofsafe water, lack of knowledge or facilities forsafe preparation and storage of food(e.g., refrigeration, fuel for hot holding or thor-ough reheating), and lack of time to prepare foodproperly before each meal. As a result, somehouseholds, particularly low-income ones, aresimply not able to apply essential food safetyprinciples, such as feeding infants with freshlyprepared foods, chill storage, hot storage, reheat-ing of stored foods, and so forth.

Thus, fermentation provides an economicmeans of preserving food and inhibiting thegrowth of pathogenic bacteria even under condi-tions where refrigeration or other means of safe

storage are not available. At the same time, it canenhance the nutritional quality of some foods. Inseveral African countries, the technology is usedin particular in the preparation of foods for in-fants and young children.2 For instance, in Kenya,Nigeria, the United Republic of Tanzania andUganda, it is customary to give infants fermentedcereals, or root-crop products. Fermentation isalso used to produce beverages. Again, in areaswhere the safety of the water cannot be ensured,fermentation processes contribute to reducing therisk of waterborne diseases.

PRACTICAL INTERVENTION TOENHANCE SAFETY OF FERMENTEDFOODS

To take advantage of the benefits that fermen-tation offers and, at the same time, to minimizeits risks, it is important to examine the fermenta-tion process carefully and to develop a planwhere hazards associated with the different pro-duction steps are considered and controlled. InChapter 3, the concept of Hazard Analysis andCritical Control Point (HACCP) as a method offood safety assurance was explained, and a planfor a fermented indigenous food that is commonin Africa and prepared at household level waspresented. Other examples of the application ofthe HACCP concept to fermented African foodshave also been published elsewhere.46"48 In thissection, a second example of an indigenous Afri-can food, togwa, is presented to illustrate howthe concept of HACCP can be used to evaluate afood production or preparation process in orderto identify possible safety problems and howthey may be controlled.

This study illustrates that the process of togwapreparation, as described here, leads to a high-risk product. The risk lies in the fact that "powerflour" (an ARF obtained from germinatedseeds), which may be contaminated, is added af-ter cooking. If fermentation fails, proliferationof contaminant bacteria may occur and acid-re-sistant pathogens may survive. Possible optionsto reduce risks include accelerated fermentationby back-slopping or use of starter culture.21 Ex-treme care should be taken to prevent

postcooking contamination. The HACCP studyof this product identifies numerous critical con-trol points (CCPs) reflecting the high degree ofconcern. The reasons for concern are that (1) thefood is intended primarily for infants and youngchildren, who, due to their susceptibility to in-fections, require great care in the preparation oftheir food; (2) a wide range of hazards have beenconsidered in this study; and (3) the preparation isenvisaged in a household setting with rudimentaryenvironmental conditions. Because the risk ofcontamination in this setting is higher, severalsteps in the preparation likely to prevent or controlcontamination need to be carefully controlled.

A second example shows an HACCP study forthe production of a traditional cheddar cheese.Cheddar cheese has been implicated in outbreaksof salmonellosis and Staphylococcus aureus in-toxication.19'25'49 In one outbreak, the problem wasdue to improper application of the HACCP system(i.e., inadequate pasteurization and a lack of cor-rective action when the monitoring results indi-cated failure in reaching critical limits).49

Both examples also underline the importanceof good hygienic practice (GHP) as a prerequi-site for the production of fermented foods, par-ticularly with regard to the prevention of con-tamination. In the case of cheddar cheese, it canbe seen that after the pasteurization step, there isno other step that can eliminate pathogens.Therefore, GHP is extremely important to pre-vent recontamination.

HACCP Study of Togwa*t

1 . Product description — Togwa is preparedby many tribes in Tanzania, and there arevariations in its production. The prepara-tion involves mixing cooked cereals (e.g.,

* Application of the HACCP system has been simpli-fied and adapted to household conditions. Although thesame approach can be used for production on a cottageand industrial scale, the requirements in terms of criticalcontrol points, critical limits, and monitoring proce-dures may be different and more severe.

fModel HACCP plans are not appropriate for useuntil validated for a specific food and food process.

maize, sorghum) with germinated cereals(e.g., finger or bulrush millet or sorghum).Fermentation takes approximately 9-12hours, depending on whether or not an oldbatch is used as a starter culture. If an oldbatch or power flour is used, the fermenta-tion is completed in approximately 6-9hours.

2. Intended use — Most fermented gruels inTanzania remain edible for one to twodays. Beyond this period, the product istoo sour, or gives off an unpleasant odor.Fermented gruels are traditionally given tochildren less than five years of age.

3. Flow diagram — Figure 12-2 shows theflow diagram of togwa.

4. Hazards of concern — Hazards consid-ered in this context include biological(e.g., bacteria, viruses, parasites), chemi-cal (e.g., contaminants, mycotoxins), andphysical agents.

5. Identification of hazards, control mea-sures, and CCPs — Table 12-2 shows haz-ards associated with each step in the prepa-ration of togwa and of power flour, which isan essential ingredient in this gruel,a. Raw material: Major hazards in maize,

millet, and sorghum are toxins (e.g.,aflatoxin produced by molds, and agro-chemicals). For prevention of the activ-ity of toxigenic molds during storage,the raw material should, as far as pos-sible, be stored under appropriate con-ditions. When the ambient temperatureand humidity are high, storage timeshould be limited. Insofar as agro-chemicals are concerned, householdscannot do much except to get assurancefrom suppliers concerning the safety ofthe products. The possibility of acci-dental contamination of grains duringstorage with agrochemicals should beprevented.

The grains may also contain foreignmatter such as stones and insect frag-ments. These will not be eliminated at alater step, so it is important that house-holds thoroughly clean the raw material.

Water used in the preparation ofslurry may be contaminated. Boilingthe slurry (see step 9) will eliminateeventual pathogens. Therefore, waterused for the preparation of slurry is nota CCP, although as part of a good hy-

gienic practice, safe water should beused as far as possible in the prepara-tion of togwa and for washing handsand utensils. On the other hand, the useof safe water in the preparation ofpower flour is essential because this

Note: Numbers correspond with those in Table 12-2.*Although this step was not part of the original togwa preparation, it was added after the HACCP study becauseboiling water was identified as being essential and needed to ensure the hygienic quality of power flour.

Figure 12-2 Flow diagram of togwa. The high number of CCPs identified in this study is due to the fact that thestudy is considered in poor hygienic conditions.

1. Sorghum, millet

2. Soaking

3. Germination

4. Sun drying

5. Storage of sun-driedgerminated grains

6. GrindingPower flour(Storage)

1. WaterBoiling andcooling*

1. Maize, sorghum,millet

7. Grinding

8. Preparation of slurry(10-15% solids)

9. Boiling

10. Cooling (toambient temperature)

11. Addition of powerflour and togwa

12. Fermentation

13. Serving togwa

Table 12-2 HACCP Study of the Preparation of Togwa in Households

Corrective Actions

a. Discard the rawmaterial and changesupplier

Utilize the rawmaterial as quickly aspossible.

d. Reclean.

a. Use another sourceof water.

b. Reboil.

Monitoring Procedure

a. Observation,smelling

Time keeping(measurement oftemperature orhumidity if possible)

d. Observation

a. Observation,smelling, and tasting

b. Observation

Critical Limit*

a. i) No moldiness,good smell

ii) Short storage time,adequate tempera-ture humidity ofstorage area

d. No visible stones

a. Clear, free of odorand off taste

b. Bubbles

CCPs

a. Yes

b. No

c. No

d. Yes

a. Yes

b. Yes forstep 2;No forstep 8

Control Measures

a. \) Obtain assurancefrom supplier ofadequate preharvestand postharvesthandling of grainsii) Store grains in dry(and if possible cool)area, limit storage time

b. Obtain assurance fromsupplier of adequatepreharvest andpostharvest handling ofgrains

c. Heat treatment,fermentation

d. Manual cleaning

a. Obtain assuranceabout the source ofwater; use only safewater

b. If safe water (i.e.,filtered and disinfected)is not available, boilthe water

Hazards

a. Mycotoxins

b. Agrochemicals

c. Pathogens:Bacillus cereus,Salmonella,Escherichia coll

d. Physical: insectsand stones

a. Chemical contami-nants, dependingon the source

b. Pathogens (e.g.,Escherichia coll,Campylobacter,V. cholerae,Salmonella,Cryptosporidium,Giardia lamblia,Entamoebahistolytica)Rotavirus

Step

1 . Raw materiali) Maize

SorghumMillet

1 . Raw materialii) Water

continues*ln this study, the preparation of togwa is considered in a household environment; therefore, the critical limits are expressed in qualitative values.

Table 12-2 continued

Corrective Actions

Refresh water.

Remove moldy grains.

a. Clean if possible. Ifnot, discard

b. As long as there is nomold growth, redryunder properconditions; otherwise,discard.

a. Clean if possible. Ifnot, discard

b. Discard

continues


Observation, smelling

Observation

a. Observation

b. Time keeping,observation

Observation

Critical Limit*

Water should remainfree from odor orfoam

Grains should remainfree of moldiness

a. No foreign matter

b. Sufficient time,adequate exposureto sun, dry ambientconditions, adequateair circulation, nomold

a. No foreign matter

b. Dry conditions ofstorage, no mold

CCPs

Yes

Yes

Yes

Yes

No

Control Measures

As far as possible at lowtemperatures

As far as possible at lowtemperature

a. Protect the sprouts

b. Ensure thorough andfast drying

a. Protect the grains

b. Keep dry

Use clean and properlymaintained equipment

Hazards

Growth of microor-ganisms

Growth of microor-ganisms (e.g.,toxigenic molds)

a. Contamination andintroduction offoreign matter

b. Inadequate dryingmay lead to growthof microorganismsduring storage

a. Contamination/introduction offoreign matter

b. Growth of toxigenicmolds, if themoisture content ishigh

Introduction of filth,dirt and foreignmatter

same as step 6

Step

2. Soaking

3. Germination

4. Sun drying

5. Storage ofsun-driedgerminatedgrains

6. Grinding topower flour

7. Grinding


Corrective Actions

Reboil.

a. Reboil.

b. Depending on thenature of contamina-tion, either clean,reboil or discard

a. & b. Use anotherpower flour or togwa.

a. Discard the material.

Reheat the foodthoroughly.


Observation

a. Time keeping

b. Observation

a. & b. Observation

a. Observation

Observation

Critical Limit*

Bubbles

a. Short time, roomtemperature withinfour hours

b. No foreign matter

a. Power flour of highhygienic quality

b. Absence of diseaseupon consumption ofpreviously preparedtogwa

a. Acid taste andcharacteristic odorwithin 24 hours

Hands properly washedwith soap and cleanwater

CCPs

No

Yes

Yes

Yes

a. Yes

b. No

Yes

Control Measures

Use clean utensils andsafe water

Thorough boiling

a. Ensure rapid cooling

b. Protect the porridgeduring the coolingprocess

a. Ensure hygienic qualityof power flour

b. Ensure the safety ofpreviously preparedtogwa

a. Rapid fermentation

b. Minimize contamina-tion with acid-tolerantpathogens (seestep 11)*

Wash hands and useclean utensils

Hazards

Contamination withpathogens throughutensils and/orwater

Survival of pathogens

a. Growth of bacterialspores

b. Contamination

a. Contamination withpathogens bypower flour

b. Contamination withacid-resistantpathogens bytogwa

a. Growth andformation of toxinby Staphylococcusaureus

b. Survival of acid-tolerant pathogens

Recontamination withpathogens byhands, utensils,and environment

Step

8. Slurrypreparation

9. Boiling

10. Cooling

11. Addition of:a. powerflour

b. Togwa

12. Fermentation

13. Serving

*As there is no critical control point in the subsequent steps that would ensure the killing of acid-tolerant pathogens surviving the fermentation step, the present process of togwapreparation may lead to a high-risk product.

process does not include a step thatwould ensure the elimination of patho-gens introduced through the raw mate-rials (i.e., water, sorghum, or millet). Inaddition, pathogens introduced throughthe raw material may proliferate duringthe soaking and germination periodsand may also survive the sun-dryingstage. Therefore, the safety of waterused in power flour preparation is criti-cal for minimizing contamination. Thehygienic quality of the power flour isparticularly important for the safety ofthe final product because there is nostep that would ensure the killing ofacid-resistant pathogens after thepower flour has been added. Therefore,after the HACCP study, it was sug-gested to add a boiling step in thepreparation of power flour to ensuresafety of water and minimize the con-tamination of power flour,

b. Soaking: During the soaking period,bacterial growth will occur. The use ofsafe water may minimize the final bac-terial load. If at all possible, soakingshould take place at low temperaturesin order to minimize the growth of mi-croorganisms,

c. Germination: This usually takes placein layers a few centimeters thick,spread on mats or leaves, and coveredwith leaves, mats, or gunny sacks toavoid excessive dehydration. Fromtime to time, the material must be airedand mixed while checking and adjust-ing the degree of grain humidity. Fur-ther contamination by mats and micro-bial growth can occur at this stage.Microbial growth, particularly with re-spect to toxigenic molds, can also oc-cur. As far as possible, germinationshould be carried out in cool condi-tions in order to minimize microbialgrowth,

d. Sun drying: During this step, the coveris removed and the sprouted grains arespread on mats or bamboo trays to dry

in the sun. This may take a few days.All kind of contaminants and foreignmatter can fall into the sprouted grainsif they are not well protected. In addi-tion, insufficient drying may lead to amicrobially unstable product duringsubsequent storage. Thorough drying iscritical for the stability of the product,

e. Storage of sun-dried germinatedgrains: Hazards associated with thisstep are contamination (e.g., by rodentsand other animals during storage) andmicrobial growth, particularly toxi-genic molds, if the germinated grainsare not properly dried or are kept in hu-mid conditions,

f. Grinding to power flour: This step mayintroduce dirt and foreign matter intothe product. As part of a good hygienicpractice, this should be avoided as faras possible by using clean and properlymaintained equipment. However, it isunlikely that this step will introduceany major health hazard,

g. Grinding: Same as step f.h. Preparation of slurry: Except for the

safety of water and cleanliness of uten-sils, no other major hazard is associatedwith this step. Because the subsequentboiling step will kill the pathogens thatmay have been introduced at this step,it is not considered a CCP. Neverthe-less, as part of a GHP, householdsshould use clean utensils and safe wateras much as possible,

i. Boiling: Boiling should be thorough inorder to gelatinize all the starch. Thisstep is also essential to kill nearly allpathogens (bacterial spores may sur-vive),

j. Cooling: The pot should be covered toprotect against dirt or other foreignmatter falling in. It is important thatcooling is carried out as fast as pos-sible. Prolonged cooling may presentan opportunity for bacterial spores togrow. When large quantities of togwaare prepared, the cooling time can be

reduced by dividing the togwa intosmall portions,

k. Addition of power flour and togwa:Two ingredients may be added here toinitiate the fermentation, that is, powerflour and previously fermented togwa.The contamination brought about bypower flour is diverse and difficult tocontrol. Viruses and other acid-tolerantagents are of particular concern here.The togwa starter culture is quiteacidic, and thus will contain only thosecontaminants that are acid tolerant.

During the fermentation that fol-lows, acid-sensitive pathogens may bekilled. However, acid-tolerant patho-gens may survive. A power flour pre-pared under hygienic conditions mayminimize the contamination of togwa.However, in the absence of a final kill-ing step such as thorough reheating,the presence of acid-tolerant pathogensin the final product may not be ex-cluded.

1. Fermentation: During fermentation, arapid dominance of LAB may be ex-pected. This is supported by the shortfermentation time to reach the requiredacidity. A rapid fermentation is criticalfor killing acid-sensitive pathogens andfor preventing bacterial growth and theproduction of toxin. The addition oftogwa enhances fermentation and maybe beneficial provided it does not intro-duce acid-resistant pathogens,

m. Serving: It is important to ensure thatpathogens are not re-introduced intotogwa by dirty hands and utensils.Therefore, these have to be washedcarefully with safe water. Dependingon the hygienic measures taken to pre-pare togwa, the final product could bemore or less contaminated becausethere is no final CCP that would ensurethe killing of acid-resistant pathogens.Thorough reheating would greatlycontribute to the safety of the final

product. However, implications interms of textural and other changesshould be considered because the finalproduct may become unacceptable tothe consumer.

HACCP Study of Traditional CheddarCheese*23

1 . Product description — Cheddar cheese isa hard-pressed cheese that is originallyfrom the United Kingdom. It has a firmbody and closed texture. It is composed ofapproximately 37% water, 33% fat, 25%protein, 1% lactose, and 4% ash. The curdis textured after cutting and scalding. It isthen milled and salted and is pressed in amold. The cheese is wrapped in firmblocks and stored at approximately4-6 0C. The cheese is stored from a fewweeks to several months, during whichtime it ripens and develops its flavor (mildcheese up to 3 months; medium cheese 6months, and mature cheese 8-12 months).Starter culture used for the production ofcheddar cheese consists of Lactococcuslactis subsp. lactis and Lc. lactis subsp.cremoris. Other organisms that also play arole during ripening are Lactobacilluscasei, Lb. plantarum, and Lb. brevis.

2. Intended use — General population, in-cluding vulnerable groups.

3. Flow diagram — Figure 12-3 shows theflow diagram of traditional cheddarcheese manufacture.

4. Hazards of concern — Major hazards areof bacterial origin. Several pathogens suchas Mycobacterium tuberculosis, Brucellaabortus, Salmonella, Campylobacter, andListeria monocytogenes may survive incheddar cheese. Other hazards of concernare antibiotic residues and histamine.

5. Identification of hazards, control mea-sures, and CCPs— Table 12-3 shows

*Model HACCP plans are not appropriate for useuntil validated for a specific food and food process.

hazards associated with each step of theproduction of cheddar cheese.• Raw material: Raw milk when re-

ceived by a dairy may be contaminatedwith a wide range of pathogens, as

mentioned above, as well as spoilageorganisms. In addition, milk may con-tain antibiotic residues. Hygienic han-dling of milk at the farm and duringtransport is essential to prevent con-

Note: Numbers correspond with those in Table 12-3.

Figure 12-3 Flow diagram of traditional cheddar cheese

1. Raw milk

2. Pasteurizing

4. Ripening

6. Setting the vat/cutting the curd

7. Heating

8. Draining and milling

9. Salting

10. Pressing

11. Packaging and aging

12. Storing and distribution

3. Starter culture

5. Rennet or rennet-like enzymes

Table 12-3 HACCP Study of the Preparation of Traditional Cheddar Cheese in Industrial Setting

Corrective Actions

b. Reject the milk.

Repasteurize.

b. Correct the tempera-ture and the amountof starter culture.

continues


b. Chemical ormicrobiological tests

Temperature, time

b. Temperaturetitratable acidity

Critical Limit

b. According to CodexAlimentariusCommission

72 0C for 1 5 seconds orequivalent

b. 30-31 0C, develop-ment of acidity of0. 14% before step 8.Proportion of starterculture and milkaccording to theinstructions of themanufacturer

CCPs

a. No

b. Yes

Yes

No

a. No

b.Yes

No

Control Measures

Hygienic collection at thefarm and transportRefrigeration (maxi-mum temperature 7 0C)

b. Control of incomingmilk

Heating

Assurance by the supplierof the starter culture onthe quality

a. Cleaning anddisinfection of vat andother equipment

b. Follow instructions ofmanufacturer of starterculture; use appropri-ate proportion ofstarter culture andmilk; keep milk warm

Assurance of quality bythe supplier

Hazards

a. Pathogenicbacteria: Mycobac-teriumspp.,Brucella spp,Campylobacter,Listeria, Salmonella

b. Antibiotic residues

Survival of pathogens

Presence of hazardsin the starterculture

a. Contamination withpathogens

b. Insufficientfermentation

Presence ofpathogens

Step

1. Rawmaterial:milk

2. Pasteuriza-tion

3. Starterculture

4. Ripening(addition ofthe starterculture)

5. Rennet orrennet I ike-enzymes


Corrective Actions

Readjust thetemperature.

Readjust the proportionof salt and curd.

Correct the temperatureof storage. Considerremoval of theproduct if the loss oftemperature controlhas exceeded severalhours.


Temperature

Amount of salt added

Temperature

Critical Limit

370C

1.7%

50C

CCPs

No

Yes

No

Yes

No

Yes

No

Control Measures

Cleaning and disinfectingof cutting knives andhygienic practice

Control temperature ofheating

Hygienic practice

Adjusting the salt leveland its distribution

Strict hygiene of theequipment andpersonnel

Adequate storage (seealso step 8)

Adequate packaging,refrigeration, and rapiddistribution

Hazards

Contamination withpathogens, ofutensils used foradding the rennet

Arresting thefermentation

Contamination ofutensils andenvironment withhazards

Growth of S.aureus atsubsequent steps

Contamination of thepress withpathogens

Growth of Staphylo-coccus aureus;growth of decar-boxylase positivelactic acid bacteriaand formation ofhistamine

Mold growth

Step

6. Setting thevat andcutting thecurd

7. Heating

8. Cheddaring(draining andmilling)

9. Salting

10. Pressing

11. Aging

12. Storing anddistributing

Note: This table was adapted from ICMSF.23

tamination and keep the level of patho-gens possibly present to a minimum.To ensure that raw material is of ad-equate hygienic quality, the milk maybe tested for its bacterial load with me-thylene blue or by some other rapidtest. The receipt of the raw milk is aCCP for antibiotic residues becausethese will not be eliminated by furtherprocessing of the milk into cheese.

• Pasteurization: The pasteurization stepis a CCP because it is essential for theelimination of pathogens that may bepresent in the raw milk. It will also re-duce the number of decarboxylase-positive LAB able to produce hista-mine.

• Starter culture: The contamination ofstarter culture has in the past causedfood-borne disease outbreaks. There-fore, the possibility that the starter cul-ture may be contaminated or be of poorquality (leading to poor fermentation)should be considered, and assuranceshould be obtained from the supplier ofits quality.

• Ripening of the milk: At this stage, themilk is cooled to 30-31 0C and pouredinto the cheese vat, and the starter cul-ture is added. It is important that thestarter culture is of adequate qualityand is added in sufficient quantity toensure rapid fermentation. The tem-perature of the milk should also be keptat an optimal level to ensure adequatestarter culture activity. Starter cultureactivity may be measured by pHchanges. The cheese vat and otherequipment should be cleaned and disin-fected before use. To prevent postpas-teurization contamination, strict hy-giene should be applied at this stageand later stages of production.

• Rennet or rennet-like enzymes: As forstarter culture, it is important to obtainassurance from the supplier that therennet is of adequate quality.

• Setting the vat and cutting the curd:The rennet or rennet-like enzymes are

added and the milk is left undisturbedto form a gel. The coagulated milk isthen cut into cubes. To prevent con-tamination with cutting utensils, hy-gienic practices are strictly observed.

• Heating: The curd is heated to 37-390C and the curd and whey are stirred.The curd shrinks as a result of heatingand the whey is released. It is importantto have control of temperature in orderto maintain the activity of the starterculture. At the end of this step, thewhey should have a titratable acidity of0.14-0. 1 6% and the curd pH should beapproximately 6.

• Draining and milling: The whey isdrained from the vat. The curd devel-ops a plastic consistency as a result ofcontinued acid development. The curdis cut into sections and passed througha mill to form ribbons. No major haz-ard is associated with this step.However, hygienic practices shouldbe strictly observed to prevent recon-tamination.

• Salting: Salt is added to shrink the curdand further separate the whey. Theamount of salt should be carefully con-trolled because it will have an impacton subsequent acid development andfavor growth of Staphylococcusaureus. The salt content should be ap-proximately 1.7%.

• Pressing: After salting, the curd isplaced in a box and pressed. In thisway, the remaining whey is removed.To prevent recontamination, this stepshould be carried out under strict hy-gienic conditions.

• Aging: The cheese is vacuum packedand aged at approximately 5 0C for afew weeks to few months. The controlof temperature at this stage is importantand critical for the prevention ofgrowth of possibly present Staph.aureus and/or activity of decarboxy-lase-positive bacteria.

• Storing and distribution: To ensuresafety and quality, storage should be at

refrigeration temperature and the prod-uct should be distributed rapidly.

RESEARCH NEEDS

Research needs in the area of fermentation areextensive and relate to several areas.

• identification of new organisms and tech-nological developments leading to newproducts with enhanced nutritional or orga-noleptic quality

• assessment of safety of starter cultures, bethey traditional organisms selected for theirspecific features or genetically modifiedorganisms

• interaction of LAB with pathogens in theintestinal tract and risks for transfer ofpathogenicity and virulence

• potential of fermentation technology tocontrol pathogens

• identification and selection of organismswith potential health benefits

• transfer of technology to settings where thetechnology can be of specific benefit tohealth

• implementation and evaluation of healtheducation interventions carried out to en-sure safe application of fermentation.

REFERENCES

1. Adams, M. R. & Nicolaides, L. (1997). Review of thesensitivity of different pathogens to fermentation. FoodControl 8, 227-239.

2. Adams, M. R. (1998). Fermented weaning foods. InM-crobiology of Fermented Foods. Vol. 2, 2nd edn., pp.790-811. Edited by B. J. B. Wood. London: BlackieAcademic and Professional.

3. Altekruse, S. F., Babagaleh, T. B., Mowbray, J. C. etal. (1998). Cheese-associated outbreaks of human ill-ness in the United States, 1973 to 1992: sanitary manu-facturing practices protect consumers J Food Prot 61,1405-1407.

4. Baldwin, L, Hobson, P., Weinstein, P. et al. (1992). Sal-monella anatum foodborne outbreak, South Australia.Commun Dis Intell 16, 94-98.

5. Bar-Dayan, Y., Bar-Dayan, Y. et al. (1996). Foodborneoutbreak of streptococcal pharyngitis in an Israeli baseair force. Scandavian J Infect Dis 28, 563-566.

Additional guidance on research needs andpriorities in this area are provided elsewhere.6'18

CONCLUSION

Fermentation is and will continue to remainan important technology. For foods for infantsand young children, it presents clear nutritionaland safety advantages. In the industrializedcountries, the technology may not be importantas a preservation technique, but advances in ar-eas such as molecular biology will help in thedevelopment of new products, with diversifiedtaste and aroma and perhaps health benefits.The challenge will be to ensure that organismsused as starter culture are safe and that the pro-cess is designed and applied in such a way thatpotential hazards, in particular acid-resistantpathogens, are controlled.

The greatest benefits but also risks are currentlyin the developing countries and in places wherethe technology is used in an artisanal way. In thedeveloping world, where refrigeration facilitiesare lacking, the technology can make a great con-tribution to public health by preventing the growthof pathogens. However, substantial health educa-tion campaigns are necessary to ensure that thetechnology is applied in an appropriate mannerand leads to safe and nutritious products.

6. Buckenhiiskes, H. J. (1997). Fermented vegetables. InFood Microbiology: Fundamentals and Frontiers. Editedby M. P. Doyle, L. Beuchat & T. J. Montville. Washing-ton, DC: American Society of Microbiology Press.

7. BGW/RKI. (1997). Abschlussbericht der Untersuch-ung eines Ausbruchs von Gastroeneritis im landkreisAnhalt-Zerbst, 1997. Internal Report from theBundesinstitut fur gesundheitlichen Verbraucherschutzund Veterindrmedizin. Berlin, Germany.

8. Caplice, E. & Fitzerald, G. F. (1999). Food fermenta-tion: role of microorganisms in food production andpreservation. IntlJ Food Microbiol 50, 131-149.

9. CCDR. (1999). Salmonella enteritidis outbreak due tocontaminated cheese: Newfoundland. Canada CommDis Report, 25-1.

1 0. CDC. (1 995). Community outbreak of hemolytic uremicsyndrome attributable to Escherichia coli O111:NM:South Australia. MMWR 550-551, 557.

11. CDC. (1995). Escherichia coli O157:H7 outbreaklinked to commercially distributed dry-cured salami:Washington and California. MMWR 44, 157-160.

12. Cowden, J. M., O'Mahony, M. et al. (1989). A nationaloutbreak of Salmonella typhimurium DT 124 caused bycontaminated salami sticks. Epidemiol and Infect 103,219-225.

13. De Roos, N. M. & Katan, M. B. (2000). Effects ofprobiotic bacteria on diarrhoea, lipid metabolism, andcarcinogenesis: a review of papers published between1988 and 1998. Am J CUn Nutr 71, 405-41 1.

14. Desenclos, J. -C., Bouvet, P. & Benz-Lemoine, E.(1996). Large outbreak of Salmonella serotypeparatyphi B infection caused by a goat's milk cheese,France, 1993: a case finding and epidemiological study.BrMedJlU, 91-94.

15. Ellis, A., Preston, M., Borczyk, A. et al. (1998). A com-munity outbreak of Salmonella berta associated with asoft cheese product. Epidemiol and Infect 120, 29-35.

16. Espinosa, F. H. et al. (1983). Group C Streptococcal in-fections associated with eating homemade cheese: NewMexico. MMWR 32, 510-516.

17. FAO/WHO. (1996). Biotechnology and Food Safety.Report of a Joint FAO/WHO Consultation, Rome Foodand Agriculture Organization of the United Nations.Rome, Italy.

18. FAOAVHO. (1996). Fermentation: Assessment and Re-search. Report of a Joint FAO/WHO Workshop on Fer-mentation as Household Technology To Improve FoodSafety. WHO Document WHO/FNU/FOS/96.1.Geneva: World Health Organization.

19. Fontaine, R. E., Mitchell, L. C. et al. (1980). Epidemicsalmonellosis from cheddar cheese: surveillance andprevention. Am J Epidemiol 111, 247-253.

20. Hauschild, A. H. W. & Gauvreau, L. (1985). Foodbornebotulism in Canada, 1971-84. Canadian Med Assoc J133,1141-1146.

21. Holzapfel, W. (1997). Use of starter cultures in fermen-tation on a household scale. Food Control 8, 241-258.

22. Hove, H., Norgaard, H. & Brobech, M. (1999). Lacticacid bacteria and the human gastrointestinal tract. Eur JClinNutr 53, 339-350.

23. ICMSF. (1988). Microorganisms in Foods 4: Applica-tion of the Hazard Analysis Critical Control Point(HACCP) System To Ensure Microbiological Safety andQuality. London: Blackwell Scientific Publications.

24. James, S. M., Fannin, S. L., Agee, B. A. et al. (1985).Listeriosis outbreak associated with Mexican-stylecheese. MMWR 34, 357-359.

25. Johnson, E., Nelson, J. & Johnson, M. (1990). Microbi-ology safety of cheese made from heat treated milk, partII: microbiology. JFoodProt 53, 519-540.

26. Khamboonruang, C. & Nateewatana, N. (1975). Trichi-nosis: a recent outbreak in Northern Thailand. SoutheastAsian J Trop Med and Public Health 6, 74-78.

27. Linnan, M. J., Mascola, L., Lou, X. D. et al. (1998). Epi-demic listeriosis associated with Mexican-style cheese.NEnglJMed3l9, 823-828.

28. Maguire, H., Cowden, J., Jacob, M. et al. (1992). Anoutbreak of Salmonella dublin infection in England andWales associated with unpasteurized cows' milk cheese.Epidemiol and Infect 109, 389-396.

29. Marier, R., Wells, J. G., Swanson, R. C. et al. (1973). Anoutbreak of enteropathogenic Escherichia colifoodborne disease traced to imported French cheese.LancetDec. 15, 1376-1378.

30. Mayer, K. & Pause, G. (1972). Biogene amines inSauerkraut. Leben.-Wiss.-Technol 5, 108-109.

3 1 . Mead, P. S. et al. (1999). Food related illness and deathin the United States. Emerging Infect Dis 5, 607-625.

32. Meng, X., Karasawa, T., Zou, K. et al. (1997). Charac-terization of a neurotoxigenic Clostridium butyricumstrain isolated from the food implicated in an outbreakof foodborne type E botulism. J CHn Microbiol 35,2160-2162.

33. MOH (1993). Clostridium perfringens food poisoningat Tabuk Camp. Saudi Epidemiol Bull 1, 2.

34. Mollet, B. (1999). Genetically improved starter strains,opportunities for the dairy industry. IntDairyJ9, 1 1-15.

35. Morgan, D., Newman, C. P., Hutchinson, D. N. et al.(1993). Verotoxin producing Escherichia coli 0 157 in-fections associated with the consumption of yoghurt.Epidemiol and Infect 111,181-187.

36. O'Mahony, M., Mitchell, E. et al. (1990). An outbreakof foodborne botulism associated with contaminated ha-zelnut yoghurt. Epidemiol and Infect 104, 389-395.

37. Pourshafie, M. R., Saifie, M., Shafiee, A. et al. (1998).An outbreak of food-borne botulism associated withcontaminated locally made cheese in Iran. Scan J InfectDis 30, 92-94.

38. Sauer, C. J., Majkowski, J., Green, S. et al. (1997).Foodborne illness outbreak associated with a semi-dryfermented sausage product. J Food Prot 60, 161 2-16 17.

39. Shaffer, N., Wainwright, R. B., Middaugh, J. P. et al.(1990). Botulism among Alaska natives: the role ofchanging food preparation and consumption practices.West J Med 153, 390-393.

40. Sharp, J. C. M. (1987). Infections associated with milkand dairy products in Europe and North America, 1980-85. Bull WHO 65, 397^106.

41. Smith, D. H., Timms, G. L. & Refai, M. (1979). Out-break of botulism in Kenyan nomads. Annals of TropMed and Parasitol 4, 2-7.

42. Stratton, J. et al. (1991). Biogenic amines in cheese andother fermented foods: a review. J Food Prot 54, 460-470.

43. Taylor, S. L. et al. (1982). Outbreak of histamine poi-soning associated with consumption of Swiss cheese. JFood Prot 45, 455-457.

44. The Pennington Group. (1997). Report on the circum-stances leading to the 1996 outbreak of infection with E.coli 0 157 in Central Scotland: the implication for foodsafety and the lessons to be learned. April, 9-11.

45. Uribarri, J., Oh, M. S. & Carroll, H. J. (1998). D-lacticacidosis: a review of clinical presentation, biochemicalfeatures, and pathophysiolocial mechanisms. Medicine72, 73-82.

46. World Association of Industrial and Technological Re-search Organizations. (1998). HACCP Sy stem for Tradi-tional African Fermented Foods: Kenkey. Taastrup,Denmark: Danish Technological Institute.

47. World Association of Industrial and Technological Re-search Organizations. (1998). HACCP System for Tradi-tional African Fermented Foods: Soumbala. Taastrup,Denmark: Danish Technological Institute.

48. World Association of Industrial and Technological Re-search Organizations. (1998). HACCP System for Tradi-tional African Fermented Foods: Soy-ogi. Taastrup,Denmark: Danish Technological Institute.

49. Wall, P. (1998). Food-borne human disease: is it a vet-erinary problem? Vet J 5 1 , 34-3 5 .

50. World Health Organization. (1991). Strategies forAssessing the Safety of Foods Produced by Biotechnol-ogy: Report of Joint FAO/WHO Consultation. Geneva:Author.

51. World Health Organization. (1996). Konzo, a distincttype of upper motoneuron disease. Weekly EpidemiologReel 1, 225-228.

52. World Health Organization. (1997). World Health Re-port. Geneva: Author.

53. Wood, G. F. & Adams, M. R. (1992). Effects of acidifi-cation, bacterial fermentation and temperature on thesurvival of rotavirus in a model weaning food. J FoodProt 55, 52-55.

54. Yusof, R. M., Morgan, J. B. & Adams, M. R. (1993).Bacteriological safety of a fermented weaning food con-taining L-lactate and nisin. J Food Prot 56, 414-417.

List of Sources

CHAPTER 1

Figure 1-2 Adapted with permission from H.Uhlig, Enzyme Arbeiten Fur Uns. TechnischeEnzyme und Ihre Anwendung, p. 37, © 1991,Carl Hanser Verlag.

Figure 1-3 Reprinted with permission fromA.G.J. Voragen and L.A.M. Van den Broek,Fruit Juices, in Biotechnological Innovations inFood Processing, Biotol Team, ed., p. 189, ©1991, Butterworth-Heinemann, Ltd.

Figure 1-4 Reprinted with permission fromN.C. Carpita and D.M. Gibeaut, Structural Mod-els of Primary Cell Walls in Flowering Plants:Consistency of Molecular Structure with thePhysical Properties of the Walls During Growth,The Plant Journal, Vol. 3, p. 7, © 1993,Blackwell Science Ltd.

Figure 1-5 Adapted from RJ. Sturgeon, Ad-vances in Macromolecular Research Vol. 1,Copyright 1997, pp. 1-64, with permission fromElsevier Science.

CHAPTER 2

Figure 2-1 Reprinted with permission fromM.R. Adams and CJ. Hall, Growth Inhibition ofFood-Borne Pathogens by Lactic and Acetic Ac-ids and Their Mixtures, InternationalJournal ofFood Science Technology, Vol. 23, pp. 287-292,© 1988.

Table 2-2 Data from A. Ouwehand, Antimicro-bial Components From Lactic Acid Bacteria, inLactic Acid Bacteria: Microbiology and Func-tional Aspects, S. Salminen and A. Von Wright,eds., pp. 139-159, © 1998, Marcel Dekker, Inc.

Table 2-3 Data from S. Saono, R.R. Hull, and B.Dhamcharee, eds., A Concise Handbook of Indig-enous Fermented Foods in the ASCA Countries, p.157, © 1986, The Government of Australia.

CHAPTER 3

Figure 3-1 Reprinted with permission from Y.Motarjemi and M. van Schothorst, HACCP:Principles and Practice, © 1999, World HealthOrganization.

Figure 3-2 Reprinted with permission from As-sessment of Fermentation as a Household Tech-nology for Improving Food Safety, A Report ofthe WHO Consultation, © 1996, World HealthOrganization.

Table 3-1 Reprinted with permission from As-sessment of Fermentation as a Household Tech-nology for Improving Food Safety, A Report ofthe WHO Consultation, © 1996, World HealthOrganization.

Appendix 3-A Reprinted with permission fromCodex Alimentarious Commission, Food Hy-giene, Basic Texts, © 1997, Joint FAO/WHOFood Standards Programme, Secretariat of theJoint FAO/WHO Food Standards Programme.

Figure 3-A Reprinted with permission from Co-dex Alimentarious Commission, Food Hygiene,Basic Texts, © 1997, Joint FAO/WHO FoodStandards Programme, Secretariat of the JointFAOAVHO Food Standards Programme.

CHAPTER 6

Figure 6-la Reprinted from International Jour-nal of Food Microbiology, Vol. 29, M.H. Silla-Santos, Biogenic Amines: Their Importance inFoods, pp. 213-231, Copyright 1996, with per-mission of Elsevier Science.

Figure 6-lb Reprinted from International Jour-nal of Food Microbiology, Vol. 29, M.H. Silla-Santos, Biogenic Amines: Their Importance inFoods, pp. 213-231, Copyright 1996, with per-mission of Elsevier Science.

Figure 6-lc Reprinted from International Jour-nal of Food Microbiology, Vol. 29, M.H. Silla-Santos, Biogenic Amines: Their Importance inFoods, pp. 213-231, Copyright 1996, with per-mission of Elsevier Science.

Figure 6-1 d Reprinted from International Jour-nal of Food Microbiology, Vol. 29, M.H. Silla-Santos, Biogenic Amines: Their Importance inFoods, pp. 213-231, Copyright 1996, with per-mission of Elsevier Science.

Table 6-1 Reprinted from International Journalof Food Microbiology, Vol. 29, M.H. Silla-Santos, Biogenic Amines: Their Importance inFoods, pp. 213-231, Copyright 1996, with per-mission of Elsevier Science.



Figure 6-3 Reprinted from International Jour-nal of Food Microbiology, Vol. 29, M.H. Silla-Santos, Biogenic Amines: Their Importance inFoods, pp. 213-231, Copyright 1996, with per-mission of Elsevier Science.



CHAPTER 7

Table 7-1 Data from T.A. Roberts, A.C. Baird-Parker, and R.B. Tompkin, eds., Microorgan-isms in Foods 5, Microbiological Specificationsof Food Pathogens, p. 513, © 1996, BlackieAcademic Professional.

CHAPTER 9

Chapter 9 © Crown Copyright 1997. Pub-lished with the permission of the Controller ofHer Britannic Majesty's Stationery Office. Theviews expressed are those of the author and donot necessarily reflect those of Her BritannicMajesty's Stationery Office or the VLA or anyother government department.

Table 9-1 © Crown Copyright.

Figure 9-1 © Crown Copyright.













CHAPTER 10

Figure 10-1 Reprinted from Enzyme andMicro-bial Technology, Vol. 26, T. Verrips, et al., pp.812-818, Copyright 2000 with permission fromElsevier Science.

Table 10-1 Courtesy of RABO Bank.

Figure 10-2 Courtesy of RABO Bank.

Figure 10-3 Reprinted from Ant. vanLeeuwenhoek, Vol. 70, 1996, pp. 299-316, Bar-riers to Application of Genetically ModifiedLactic Acid Bacteria, C.T. Verrips and Van denBerg, with kind permission of Kluwer AcademicPublishers.

Figure 10-6 Reprinted with permission fromFAOSTAT, C.C. Mann, Crop Scientists Seek aNew Revolution, Science, Vol. 283, pp. 310-314. Copyright 1999 American Association forthe Advancement of Science.

Figure 10-7 Reprinted from Ant. van Leeuwen-hoek, Vol. 70, 1996, pp. 299-316, Barriers to Ap-plication of Genetically Modified Lactic AcidBacteria, C.T. Verrips and Van den Berg, withkind permission of Kluwer Academic Publishers.

Figure 10-10 Reprinted from Ant. van Leeuwen-hoek, Vol. 70, 1996, pp. 299-316, Barriers to Ap-plication of Genetically Modified Lactic AcidBacteria, C.T. Verrips and Van den Berg, withkind permission of Kluwer Academic Publishers.

CHAPTER 11

Exhibit 11-2 Data from S. Salminen and A. VonWright, Current Probiotics-Safety Assured? Mi-crobial Ecology in Health and Disease, Vol. 10,pp. 68-77, © 1998 and S. Salimen, et al., Dem-onstration of Safety of Probiotics: A Review, In-ternational Journal of Food Microbiology, Vol.44, pp. 93-106, ©1998.

Table 11-1 Data from W.H. Holzapfel, et al.,Taxonomy and Important Features of ProbioticMicroorganisms in Food and Nutrition, Ameri-can Journal of Clinical Nutrition, in press and T.Mitsuoka, Intestinal Flora and Ageing, NutritionReview, Vol. 50, pp. 438-446, © 1992.

CHAPTER 12

Chapter 12 Portions of this chapter are adaptedfrom Fermentation: Assessment and Research.Report of a Joint FAO/WHO Workshop on Fer-mentation as a Household Technology to Im-prove Food Safety, Pretoria, South Africa, De-cember 11-15, 1995, © 1996, World HealthOrganization.

Figure 12-1 Courtesy of the World Health Orga-nization, Geneva, Switzerland.


Table 12-2 Courtesy of the World Health Orga-nization, Geneva, Switzerland.


Table 12-3 Data from ICMSF, Microorganismsin Foods 4: Application of the Hazard AnalysisCritical Control Point (HACCP) System to En-sure Microbiological Safety and Quality, ©1988, Blackwell Scientific Publications.

279 This page has been reformatted by Knovel to provide easier navigation.

Index

Index terms Links

A Acetic acid, Escherichia coli 43

Adenovirus 161 165

Aeromonas hydrophila 143 pH 142 temperature 142 water activity 142

Aflatoxicol 105 106

Aflatoxin 102

Aflatoxin B1 carcinogenesis 103 105 relationship between metabolism and toxicology 104 structure 103

Aflatoxin B2a 105 106

Algal toxin 113

Alkaloid 71

Almond 79

Alternaria toxin 111

Amoebiosis 208 distribution 203 209 epidemiology 209 life cycle 209 prevention and control 209 public health significance 209 transmission 209

Amplified fragment length polymorphism probiotic strain 245 starter strain 245

Amylolytic enzyme 8

280 Index terms Links


Angiostrongylosis 176 178 distribution 176 179 epidemiology 176 life cycle 176 178 prevention and control 176 public health significance 176 transmission 176

Animal origin 4

Anisakiosis 178 distribution 179 180 epidemiology 180 life cycle 180 181 prevention and control 180 public health significance 180 transmission 180

Antibiotic resistance marker, genetic modification 219 227

Antinutritional, defined 71

Apricot 79

Arabinans, structure 9

Astrovirus 161 166

B Bacillus cereus 145

pH 142 temperature 142 water activity 142

Bacteria 141 cleaning 150 contaminated raw materials 149 control of microbial hazards 148 disinfection 150 effect of fermentation processes on survival 147 function in fermentation 7 hazard analysis critical control point 150 limits for growth 142



Bacteria (Continued) pathogens of concern in fermented products 142 presence of pathogens in raw materials 146 quantitative risk assessment 150 zoning 149

Bacterial toxin 112

Bacteriocin classification 45 lactic acid bacteria 45

Balantidiosis 210 distribution 203 210 epidemiology 210 life cycle 210 prevention and control 211 public health significance 210 transmission 210

Bean 79

Beer biogenic amine 124 nitrosamine 133

Biogenic amine 119 beer 124 cheese 124 dairy product 124 factors affecting synthesis 122 in fermented food 122 fish 123 formation in food 119 health effects 125 histamine 125 meat 124 microorganisms producing 121 precursors 119 120 produced during processing 119 vegetable 124



Biogenic amine (Continued) wine 124

Bioprocessing operations 16

Biotechnology, food 239

Biotin 5

Blastocystosis 209 distribution 203 210 epidemiology 210 life cycle 209 prevention and control 210 public health significance 209 transmission 210

Bongkrek poisoning 112

Bongkrekic acid 112

C Calcium 5

Calicivirus 161

Camembert surface-ripened cheese flow diagram 38 process condition 38

Campylobacter 143 pH 142 temperature 142 water activity 142

Capillariosis 181 distribution 179 182 epidemiology 182 life cycle 182 prevention and control 182 public health significance 181 transmission 182

Carbon 5

Carbon dioxide, lactic acid bacteria 46



Carcinogenesis, aflatoxin B1 103 105

Cassava root flow diagram 21 22 household processing 84 industrial processing 85 process condition 21 22

Cellulase 12

Cereal beer flow diagram 23 process condition 23

Cereal bread flow diagram 25 process condition 25

Cestode 187 distribution 179 main hosts 179 source of infection 179

Cheddar cheese, hazard analysis critical control point 266

Cheese biogenic amine 124 genetic modification 222 safety evaluation 232 hazard analysis critical control point 266 histamine 120

Chemical hazard endogenous compound 71 toxin 101

Cleaning 15 bacteria 150 safety 40

Clonorchiosis 192 distribution 192 194 epidemiology 192 life cycle 192 193 prevention and control 192



Clonorchiosis (Continued) public health significance 192 transmission 192

Clostridium botulinum 145 pH 142 temperature 142 water activity 142

Codes of Good Manufacturing Practice 54

Codes of Hygiene Practice 54

Codex Alimentarius Commission GATT Uruguay Round of Multilateral Trade Negotiations 57 hazard analysis critical control point 53 World Trade Organization Agreement on Sanitary and Phytosanitary Measures 57

Commercial food enzyme, recombinant DNA technology 221

Commercial sterility 16

Composition, fermented fish product 48 49

Convicine 76 characterized 76 risks 82

Cooling 16

Cryptosporidiosis 200 distribution 202 203 epidemiology 202 life cycle 202 prevention and control 204 public health significance 200 transmission 202

Cucurbitacin 77 characterized 77

Cured food, nitrosamine 131

Cyanobacterial toxin 113

Cyanogenic glycoside 72 acute intoxication 78 characterized 72



Cyanogenic glycoside (Continued) chronic intoxication syndromes 78 risks 77

Cyanoginosin 113 114

Cyclosporiosis 204 distribution 203 204 epidemiology 205 life cycle 204 prevention and control 205 public health significance 204 transmission 205

Cysticercosis 188 distribution 179 190 epidemiology 190 life cycle 190 191 prevention and control 190 public health significance 188 transmission 190

D Dairy product, biogenic amine 124

Degradation enzyme 87 microorganism 87

Degradation products 86

Deoxynivalenol 109

Diacetyl, lactic acid bacteria 47

Dioctophymosis 182 distribution 179 182 epidemiology 182 life cycle 182 prevention and control 183 public health significance 182 transmission 182



Diphyllobothriosis 187 distribution 179 188 epidemiology 188 life cycle 188 189 prevention and control 188 public health significance 188 transmission 188

Disinfection, bacteria 150

E Echinostomosis 193

distribution 194 195 epidemiology 195 life cycle 195 prevention and control 195 public health significance 195 196 transmission 195

Endogenous compound 71

Enzootic hematuria 76

Enzyme 8 degradation 87 degrading cell wall components 9 genetic modification

risks 228 safety evaluation 229 232

Escherichia coli 144 acetic acid 43 lactic acid 43 pH 142 temperature 142 water activity 142

Escherichia coli O157:H7 144

Ethanol, lactic acid bacteria 47

Ethyl carbamate 126 applications 128



Ethyl carbamate (Continued) factors affecting synthesis 128 formation 127 health effects 129

Ethyl carbamide 126 produced during processing 119

Eukaryote algal toxicosis 114

F Fasciolopsiosis 197

distribution 194 197 epidemiology 197 life cycle 197 prevention and control 197 public health significance 196 197 transmission 197

Fasciolosis 195 distribution 194 195 epidemiology 197 life cycle 195 prevention and control 197 public health significance 195 transmission 197

Fermentation applications 253 factors 255 food fermentation components 4

enzyme 8 food ingredients 4 microorganism 6

importance in public health 257 industrialized societies 1 2 pitfalls 255 processes 14 tropical developing regions 1 2



Fermented fish product composition 48 49 shelf life 48 49

Fermented food diversity 13 food safety

practical intervention 259 supply chain 221

food-borne disease 255 genetic modification

functional microorganisms 224 now in market 224

research needs 271

Fermented sausage fermentation process validation 152 microbiological challenge test 153

carrying out 154 Escherichia coli O157:H7 154 planning 153 purpose 153

simplified identification procedure for food-borne microbial hazards 151

Fish 5 biogenic amine 123 flow diagram 35 36 process diagram 35 36

Fish sauce flow diagram 36 process diagram 36

Fish-rice paste flow diagram 35 process diagram 35

Flaxseed 80

Fluke. See Trematode

Folic acid 5

Food, biotechnology 239



Food contamination, sources 258

Food group 4

Food hygiene defined 53 hazard analysis critical control point 57 history 54

Food safety defined 53 fermented food

practical intervention 259 supply chain 221

history 54 reasons for concerns 54 research needs 271

Food-borne disease fermented food 255 outbreaks related to fermented products 141 precursors 119 120

Food-borne virus 160 culture 160 diagnosis 160 features 160

Fruit wine flow diagram 31 32 process condition 31 32

Fumonisin 110

Fusarium toxin 108

G Gari

flow diagram 21 hazard analysis critical control point 60

bagging 65 cooling 65 fermentation 65



Gari (Continued) flow diagram 60 61 grating 65 hazards of concern 60 intended use 60 peeling 60 product description 60 raw material 60 roasting 65 serving 65 storing 65 washing 61

process condition 21

Gastrointestinal tract, probiotic strain 239

GATT Uruguay Round of Multilateral Trade Negotiations, Codex Alimentarius Commission 57

Genetic modification 219 acceptance 222 antibiotic resistance marker 219 227 benefits 220 223

229 cheese 222

safety evaluation 232 enzyme

risks 228 safety evaluation 229 232

fermented food product functional microorganisms 224 now in market 224

lactic acid bacteria safety evaluation 232 structured assessment risk 234

microorganism benefits 228 consumer safety determination 230



Genetic modification (Continued) environmental consequences determination 230 risks 228 safety evaluation 229 232 schematic benefit vs. risk ratio 225

perception 219 plant, schematic benefit vs. risk ratio 226 probiotic strain 246 risk 219 risk assessment parameters 223 risks 227 soy sauce 222

safety evaluation 229 soybean, safety evaluation 229 starter strain 246 wheat, safety evaluation 229

Giardiosis 207 distribution 203 208 epidemiology 208 life cycle 208 prevention and control 208 public health significance 208 transmission 208

Glucosinolate 77 characterized 77 risks 82

Glycoalkaloid 75 characterized 75

Glycoside 71 antinutritional 73 bitter-tasting 73 characterized 71 structure 72 toxic 73 toxin concentration variation among varieties and cultivars 83



Gnathostomosis 183 distribution 179 183 epidemiology 183 life cycle 183 184 prevention and control 183 public health significance 183 transmission 183

Gongylonemosis 183 distribution 179 185 epidemiology 185 life cycle 185 prevention and control 185 public health significance 184 transmission 185

Grading 15 safety 39

Growth factor 5

H Hazard analysis critical control point 53

application areas 56 application guidelines 67 bacteria 150 benefits 56 characterized 55 cheese 266 Codex Alimentarius Commission 53 corrective action 70 critical control point monitoring 69 critical limit 68 determining critical control points 68 documentation 70 food hygiene 57 food preparation 58

consumer susceptibility 59



Hazard analysis critical control point (Continued) handling 59 intrinsic properties of food 58 volume of food prepared 59

gari 60 bagging 65 cooling 65 fermentation 65 flow diagram 60 61 grating 65 hazards of concern 60 intended use 60 peeling 60 product description 60 raw material 60 roasting 65 serving 65 storing 65 washing 61

hazard analysis 68 health education 59 historical development 53 management commitment 55 prerequisites 55 principles 67 togwa 259 training 55 59 verification 70 World Health Organization 53

Hazard identification procedure, fermented sausage 151

Health education, hazard analysis critical control point 59

Heating 16

Helminth classification 177 distribution 179 194



Helminth (Continued) main hosts 179 194 source of infection 179 194

Hemicellulase 9

Hepatitis virus 162 167

Heterofermenter 42

Heterophyosis 197

Histamine biogenic amine 125 cheese 120

Homofermenter 42

Household processing, cassava root 84

Hydrogen peroxide, lactic acid bacteria 46

I Industrial processing, cassava root 85

Inositol 5

Iron 5

K Koikuchi-shoyu

flow diagram 27 process condition 27

L Lactic acid, Escherichia coli 43

Lactic acid bacteria 1 antimicrobial factors associated with 43 bacteriocin 45 benefits of lactic acid fermentation 254 carbon dioxide 46 characterized 41 diacetyl 47 ethanol 47



Lactic acid bacteria (Continued) genetic modification

safety evaluation 232 structured assessment risk 234

hydrogen peroxide 46 low molecular weight compound 47 low pH 42 nutrient depletion 48 organic acid 42 overcrowding 48 principal genera associated with food 42 reuterin 47 virulence factor 240

Lactobacillus species 242

Lager beer flow diagram 23 process condition 23

Legume flow diagram 26 27 process condition 26 27

Linseed 80

Linum usitatissimum 80

Lipase 12

Listeria monocytogenes 146 pH 142 temperature 142 water activity 142

Low molecular weight compound, lactic acid bacteria 47

Lup Cheong flow diagram 34 process condition 34

M Magnesium 5

Malt alkaloid 133 134



Manganese 5

Manihot esculenta 80

Meat 4 biogenic amine 124 flow diagram 33 34 process condition 33 34

Methylazoxymethanol glycoside 75 characterized 75 risks 81

Microbial activity, safety 41

Microbial cell mass 1

Microbial enzyme 1

Microbial toxin 101

Microbiological challenge test, fermented sausage 153 carrying out 154 Escherichia coli O157:H7 154 planning 153 purpose 153

Microbiological hazard bacteria 141 parasite 175 virus 159

Microorganism 6 added food ingredients 5 degradation 87 examples 7 function in fermentation 7 genetic modification

benefits 228 determining environmental consequences 230 general scheme to determine safety to consumers and some environmental

consequences 230 risks 228 safety evaluation 229 232



Microorganism (Continued) schematic benefit vs. risk ratio 225

natural fermentation in raw substrate 6 nutrients required by 5

Milk 4 flow diagram 37 process condition 37

Milk cheese flow diagram 38 process condition 38

Mineral 5

Mixed starter culture preheated substrate 6 raw substrate 6

Mixing 16 safety 41

Moisture adjustment 15 safety 40

Mold, function in fermentation 7

N Nanophyetiosis 198

Nematode 175 distribution 179 main hosts 179 source of infection 179

Nicotinic acid 5

Nisin 45

Nitrogen 5

Nitrosamine 129 beer 133 cured food 131 formation 130 health effects 133



Nitrosamine (Continued) produced during processing 119 tobacco smoke 134

Nomenclature probiotic strain 242 starter strain 242

Norwalk-like virus 161 163

Nuoc-mam flow diagram 36 process diagram 36

Nutrient depletion, lactic acid bacteria 48

O Ochratoxin A 105

Okadaic acid 115

Oligosaccharide 76 antinutritional 73 bitter-tasting 73 characterized 76 toxic 73

Olive flow diagram 30 process condition 30

Opisthorchiosis 198

Organic acid, lactic acid bacteria 42

P Palm wine


Pantothenic acid 5

Paragonimosis 199 distribution 194 200 epidemiology 200



Paragonimosis (Continued) life cycle 200 201 prevention and control 200 public health significance 196 199 transmission 200

Parasite 175

Pasteurization 154

Patulin 107

Pectolytic enzyme 9

pH, lactic acid bacteria 42

Phaseolus lunatus 79

Physical processing, safety 39

Physical separation 15 safety 40

Phytase 12

Phytic acid 12 structure 13

Plaa-raa flow diagram 35 process diagram 35

Plant, genetic modification, schematic benefit vs. risk ratio 226

Plant cell wall middle lamella 9 11 primary cell wall 9 11 secondary cell wall 9 11 structure 9 11

Plant origin 4

Potassium 5

Power flour 60

Preheated substrate mixed starter culture 6 pure culture 6

Probiotic strain amplified fragment length polymorphism 245



Probiotic strain (Continued) gastrointestinal tract 239 genetic modification 246 host 242 nomenclature 242 randomly amplified polymorphic DNA 245 restriction enzyme analysis 245 safety assessment 239

procedures 242 243 techniques 240

taxonomic basis 242 transfer of antibiotic resistance 246

Process conditions 17

Process unit operations 15 bioprocessing operations 16 flow diagram 17 physical operations 15 thermal processing operations 16

Protease 9 10

Protozoa 200 classification 202 distribution 203 main hosts 203 source of infection to man 203

Prunus species 79

Ptaquiloside 76 characterized 76 risks 82

Pure culture preheated substrate 6 sterilized substrate 8

Pyridoxine 5

Q Quantitative risk assessment, bacteria 150



R Randomly amplified polymorphic DNA

probiotic strain 245 starter strain 245

Raw substrate, mixed starter culture 6

Recombinant DNA technology 219 commercial food enzyme 221

Red grape wine flow diagram 31 process condition 31

Restriction enzyme analysis probiotic strain 245 starter strain 245

Reuterin, lactic acid bacteria 47

Riboflavin 5

Risk assessment, genetic modification 223

Rotavirus 161 165

S Safety

antimicrobial factor significance 48 classification of food hazards 39 40 cleaning 40 examples of threats 3 grading 39 microbial activity 41 mixing 41 moisture adjustment 40 physical processing 39 physical separation 40 size reduction 41 sorting 39 transport 39 washing 40



Salami flow diagram 33 process condition 33

Salmonella 144 pH 142 temperature 142 water activity 142

Saponin 76 characterized 76 risks 82

Sapporo-like virus 161 164

Sarcocystosis 205 distribution 203 206 epidemiology 206 life cycle 205 prevention and control 206 public health significance 205 transmission 206

Sauerkraut flow diagram 28 process condition 28

Sausage raw


raw fermented flow diagram 33 process condition 33

Saxitoxin 115

Seafood 5

Shelf life, fermented fish product 48 49

Shigella 144 pH 142 temperature 142 water activity 142



Size reduction 15 safety 41

Sodium 5

Sorghum species 81

Sorting 15 safety 39

Sourdough bread flow diagram 25 process condition 25

Soy sauce flow diagram 27 genetic modification 222

safety evaluation 229 process condition 27

Soybean, genetic modification, safety evaluation 229

Soybean tempeh flow diagram 26 process condition 26

Staphylococcus aureus 146 pH 142 temperature 142 water activity 142

Starchy roots and tubers flow diagram 21 22 process condition 21 22

Starter strain amplified fragment length polymorphism 245 genetic modification 246 nomenclature 242 randomly amplified polymorphic DNA 245 restriction enzyme analysis 245 safety assessment 239

procedures 242 243 techniques 240

taxonomic basis 242



Starter strain (Continued) transfer of antibiotic resistance 246

Sterilization 16

Sterilized substrate, pure culture 8

T T-2 toxin 109

Taeniosis 188 distribution 179 190 epidemiology 190 life cycle 190 191 prevention and control 190 public health significance 188 transmission 190

Tannase 13

Tannin 71

Tape ketella flow diagram 22 process condition 22

Tapeworm. See Cestode

Tempe kedele flow diagram 26 process condition 26

Tenuazonic acid 111

Thermal processing operations 16

Thiamine 5

Tobacco smoke, nitrosamine 134

Toddy wine flow diagram 32 process condition 32

Togwa 60 hazard analysis critical control point 259

Toxic nitrogen compound, produced during processing 119



Toxin 101 removal by processing

degradation products 86 enzymes 87 microorganisms 87 modern industrial processing 84 traditional household 84 unit operations influencing 85

Toxoflavin 112

Toxoplasmosis 206 distribution 203 207 epidemiology 207 life cycle 206 prevention and control 207 public health significance 206 transmission 207

Training, hazard analysis critical control point 55 59

Transport 15 safety 39

Trematode 192 distribution 194 main hosts 194 source of infection 194

Trichinellosis (trichinosis) 185 distribution 179 186 epidemiology 187 life cycle 185 prevention and control 187 public health significance 185 transmission 187

Trichothecene 109

V Vegetable

biogenic amine 124



Vegetable (Continued) flow diagram 28 30 process condition 28 30

Vibrio cholerae 143 pH 142 temperature 142 water activity 142

Vibrio parahemolyticus 143 pH 142 temperature 142 water activity 142

Vibrio vulnificus 144

Vicine 76 characterized 76 risks 82

Virulence factor, lactic acid bacteria 240

Virus 159 contamination of foods 168 risk assessment 168 stability of virus particle 159 virus survival in fermented foods 169 water 159

Vitamin 5

W Washing 15

safety 40

Water 5 virus 159

Water activity 5

Wheat, genetic modification, safety evaluation 229

Wheat mixed grain sourdough bread flow diagram 25 process condition 25



Wine, biogenic amine 124

World Health Organization, hazard analysis critical control point 53

World Trade Organization Agreement on Sanitary and Phytosanitary Measures, Codex Alimentarius Commission 57

Y Yeast, function in fermentation 7

Yersinia enterocolitica 144 145 pH 142 temperature 142 water activity 142

Yogurt flow diagram 37 process condition 37

Z Zearalenone 108

Zoning, bacteria 149